Skip to main content

Full text of "Manuel De Landa A New Philosophy Of Society Assemblage Theory And Social Complexity"

See other formats

80 Maiden Lane 
Suite 704 

New York, NY 10038 

The Tower Building 
11 York Road 
London, SE! 7NX 


tr; Manuel DeLanda 2006 

All rights reserved. No pan of this publication may be reproduced or transmitted in 
any form or by any means, electronic or mechanical, including photocopying, 
recording, or any information storage or retrieval system, without prior permission 
in writing from the publishers. 

Manuel DeLanda has asserted his right under the Copyright, Designs and Patents Act, 1988, 
to be identified as Author of this work. 

British Library Cataloguing-in-Publication Data 
A catalogue record for this book is available front the British Library 

ISBN 0-8264-8170-1 (hardback) 0-8264-9169-3 (paperback) 

Library of Congress Cataloging-in-Publication Data 
A catalog record for this book is available from the Library of Congress 

Typeset by BookEns Ltd, Royston, Herts. 

Printed and bound in Great Britain by Ashford Colour Press Ltd, 
Gosport. Hampshire 


Introduction 1 

1 Assemblages against Totalities 8 

2 Assemblages against Essences 26 

3 Persons and Networks 47 

4 Organizations and Governments 68 

5 Cities and Nations 94 

Index 141 



The purpose of this book is to introduce a novel approach to social 
ontology. Like any other ontological investigation it concerns itself with 
the question of what kinds of entities we can legitimately commit 
ourselves to assert exist. The ontological stance taken here has 
traditionally been labelled 'realist': a stance usually defined by a 
commitment to the mind-independent existence of reality. In the case 
of social ontology, however, this definition must be qualified because 
most social entities, from small communities to large nation-states, would 
disappear altogether if human minds ceased to exist. In this sense social 
entities are clearly not mind-independent. Hence, a realist approach to 
social ontology must assert the autonomy of social entities from the 
conceptions we have of them. To say that social entities have a reality that 
is conception-independent is simply to assert that the theories, models 
and classifications we use to study them may be objectively wrong, that 
is, that they may fail to capture the real history and internal dynamics of 
those entities. 

There are, however, important cases in which the very models and 
classifications social scientists use affect the behaviour of the entities 
being studied. Political or medical classifications using categories like 
'female refugee' or 'hyperactive child', for example, may interact with the 
people being classified if they become aware of the fact that they are 
being so classified. In the first case, a woman fleeing terrible conditions in 
her home country may become aware of the criteria to classify 'female 
refugees' used by the country to which she wants to emigrate, and 
change her behaviour to fit that criteria. In this case, an ontological 




commitment to the referent of the term 'female refugee’ would be hard to 
maintain, since the very use of the term may be creating its own 
referents. On the other hand, accepting that the referents of some general 
terms may in fact be moving targets does not undermine social realism: to 
explain the case of the female refugee one has to invoke, in addition to 
her awareness of the meaning of the term 'female refugee', the objective 
existence of a whole set of institutional organizations (courts, immigra¬ 
tion agencies, airports and seaports, detention centres), institutional 
norms and objects (laws, binding court decisions, passports) and 
institutional practices (confining, monitoring, interrogating), forming 
the context in which the interactions between categories and their 
referents take place. In other words, the problem for a realist social 
ontology arises here not because the meanings of all general terms shape 
the very perception that social scientists have of their referents, creating a 
vicious circle, but only in some special cases and in the context of 
institutions and practices that are not reducible to meanings. As the 
philosopher Ian Hacking writes: 

I do not necessarily mean that hyperactive children, as individuals, on 
their own, become aware of how they are classified, and thus react to 
the classification. Of course they may, but the interaction occurs in the 
larger matrix of institutions and practices surrounding this classifica¬ 
tion. There was a time when children described as hyperactive were 
placed in 'stirn-free' classrooms: classrooms in which stimuli were 
minimized, so that children would have no occasion for excess 
activity. Desks were far apart. The walls had no decoration. The 
windows were curtained. The teacher wore a plain black dress with no 
ornaments. The walls were designed for minimum noise reflection. 
The classification hyperactive did not interact with the children simply 
because individual children had heard the word and changed 
accordingly. It interacted with those who were so described in 
institutions and practices that were predicated upon classifying 
children that way. 1 

In short, acknowledging the existence of troublesome cases in which 
the meanings of words affect their own referents in no way compromises 
a realist approach to institutions and practices. On the contrary, a correct 
solution to this problem seems to demand an ontology in which the 
existence of institutional organizations, interpersonal networks and many 

other social entities is treated as conception-independent. This realist 
solution is diametrically opposed to the idealist one espoused by 
phenomenologically influenced sociologists, the so-called 'social con¬ 
structivists'. In fact, as Hacking points out, these sociologists use the term 
'construction' in a purely metaphorical sense, ignoring 'its literal 
meaning, that of building or assembling from parts'. 2 By contrast, the 
realist social ontology to be defended in this book is all about objective 
processes of assembly: a wide range of social entities, from persons to 
nation-states, will be treated as assemblages constructed through very 
specific historical processes, processes in which language plays an 
important but not a constitutive role. 

A theory of assemblages, and of the processes that create and stabilize 
their historical identity, was created by the philosopher Gilles Deleuze in 
the last decades of the twentieth century. This theory was meant to apply 
to a wide variety of wholes constructed from heterogeneous parts. 
Entities ranging from atoms and molecules to biological organisms, 
species and ecosystems may be usefully treated as assemblages and 
therefore as entities that are products of historical processes. This implies, 
of course, that one uses the term 'historical' to include cosmological and 
evolutionary history, not only human history. Assemblage theory may 
also be applied to social entities, but the very fact that it cuts across the 
nature-culture divide is evidence of its realist credentials. It may be 
objected, however, that the relatively few pages dedicated to assemblage 
theory in the work of Deleuze (much of it in partnership with Felix 
Guattari) hardly amount to a fully-fledged theory. 3 And this is, in fact, 
correct. But the concepts used to specify the characteristics of assemblages 
in those few pages (concepts such as 'expression' or 'territorialization ) 
are highly elaborated and connected to yet other concepts throughout 
Deleuze's work. Taking into account the entire network of ideas within 
which the concept of 'assemblage' performs its conceptual duties, we do 
have at least the rudiments of a theory. But this, in turn, raises another 
difficulty. The definitions of the concepts used to characterize assemblages 
are dispersed throughout Deleuze's work: part of a definition may be in 
one book, extended somewhere else, and qualified later in some obscure 
essay. Even in those cases where conceptual definitions are easy to locate, 
they are usually not given in a style that allows for a straightforward 
interpretation. This would seem to condemn a book on assemblage 
theory to spend most of its pages doing hermeneutics. 

To sidestep this difficulty I have elsewhere reconstructed the whole of 



Deleuzian ontology, including those parts that bear directly on 
assemblage theory, in a clear, analytic style that makes a preoccupation 
with what Deleuze 'really meant' almost completely unnecessary. 4 In this 
book I will make use of a similar strategy: I will give my own definitions 
of the technical terms, use my own arguments to justify them, and use 
entirely different theoretical resources to develop them. This manceuvre 
will not completely eliminate the need to engage in Deleuzian 
hermeneutics but it will allow me to confine that part of the job to 
footnotes. Readers who feel that the theory developed here is not strictly 
speaking Deleuze's own are welcome to call it 'neo-assemblage theory', 
'assemblage theory 2.0', or some other name. 

The first two chapters of this book introduce the fundamental ideas of 
such a reconstructed theory of assemblages. This theory must, first of all, 
account for the synthesis of the properties of a whole not reducible to its 
parts. In this synthetic function assemblage theory has rivals that are 
historically much older, such as Hegelian dialectics. Thus, an important 
task, one to be carried out in Chapter 1, is to contrast assemblages and 
Hegelian totalities. The main difference is that in assemblage theory the 
fact that a whole possess synthetic or emergent properties does not 
preclude the possibility of analysis. In other words, unlike organic 
totalities, the parts of an assemblage do not form a seamless whole. In 
Chapter 2 I will argue that once historical processes are used to explain 
the synthesis of inorganic, organic and social assemblages there is no need 
for essentialism to account for their enduring identities. This allows 
assemblage theory to avoid one of the main shortcomings of other forms 
of social realism: an ontological commitment to the existence of essences. 

Once the basic ideas have been laid out, the next three chapters apply 
the assemblage approach to a concrete case-study: the problem of the link 
between the micro- and the macro-levels of social reality. Traditionally, 
this problem has been framed in reductionist terms. Reductionism in 
social science is often illustrated with the methodological individualism 
characteristic of microeconomics, in which all that matters are rational 
decisions made by individual persons in isolation from one another. But 
the phenomenological individualism of social constructivism is also 
reductionist even though its conception of the micro-level is not based on 
individual rationality but on the routines and categories that structure 
individual experience. In neither one of these individualisms is there a 
denial that there exists, in addition to rationality or experience, 
something like 'society as a whole'. But such an entity is conceptualized 

as a mere aggregate, that is, as a whole without properties that are more 
than the sum of its parts. For this reason we may refer to these solutions 
to the micro-macro problem as 'micro-reductionist'. 

The other position that has been historically adopted towards the 
micro-macro problem is that social structure is what really exists, 
individual persons being mere products of the society in which they are 
born. The young Durkheim, the older Marx, and functionalists such as 
Talcott Parsons are examples of this stance. These authors do not deny the 
existence of individual persons but assume that once they have been 
socialized by the family and the school, they have so internalized the 
values of the societies or the social classes to which they belong that their 
allegiance to a given social order may be taken for granted. This tends to 
make the micro-level a mere epiphenomenon and for this reason this 
stance may be labelled 'macro-reductionist'. There are many other 
positions taken in social science towards the problem of the articulation of 
the micro and the macro, including making an intermediate level, such as 
praxis, the true core of social reality, with both individual agency and 
social structure being byproducts of this fundamental level. This seems to 
be the stance taken by such prominent contemporary sociologists as 
Anthony Giddens, a stance that may be labelled 'meso-reductionist'. 5 

These three reductionist positions do not, of course, exhaust the 
possibilities. There are many social scientists whose work focuses on social 
entities that are neither micro nor macro: F.rving Coffman's work on 
conversations and other social encounters; Max Weber's work on 
institutional organizations; Charles Tilly's work on social justice move¬ 
ments; not to mention the large number of sociologists working on the 
theory of social networks, or the geographers studying cities and regions. 
What the work of these authors reveals is a large number of intermediate 
levels between the micro and the macro, the ontological status of which 
has not been properly conceptualized. Assemblage theory can provide the 
framework in which the contributions of these and other authors 
(including the work of those holding reductionist stances) may be 
properly located and the connections between them fully elucidated. This 
is because assemblages, being wholes whose properties emerge from the 
interactions between parts, can be used to model any of these 
intermediate entities: interpersonal networks and institutional organiza¬ 
tions are assemblages of people; social justice movements are assemblages 
of several networked communities; central governments are assemblages 
of several organizations; cities are assemblages of people, networks. 



organizations, as well as of a variety of infrastructural components, from 
buildings and streets to conduits for matter and energy flows; nation¬ 
states are assemblages of cities, the geographical regions organized by 
cities, and the provinces that several such regions form. 

Chapters 3, 4 and 5 take the reader on a journey that, starting at the 
personal (and even subpersonal) scale, climbs up one scale at a time all the 
way to territorial states and beyond. It is only by experiencing this upward 
movement, the movement that in reality generates all these emergent 
wholes, that a reader can get a sense of the irreducible social complexity 
characterizing the contemporary world. This does not imply that the 
ontological scheme proposed here is not applicable to simpler or older 
societies: it can be used in truncated form to apply it to societies without 
cities or large central governments, for example. I make, on the other hand, 
no effort to be multicultural: all my examples come from either Europe or 
the USA. This simply reflects my belief that some of the properties of social 
assemblages, such as interpersonal networks or institutional organizations, 
remain approximately invariant across different cultures. But even the! 
illustrations from Western nations are often sketchy and, with the) 
exception of Chapter 5, the historical aspects of my examples are not fully 1 
explored. This shortcoming is justified by the fact that my older publications] 
have already engaged history and historical dynamics, and that in this book 
I am exclusively interested in a clarification of the ontological status of the 
entities that are the actors of my earlier historical narratives. 6 The shortage' 
of historical examples is also intended to reduce the time the reader spends: 
at each level of scale, that is, to increase the speed of the upward 
movement, since for this book it is the reader's experience of the journey 
from the micro to the macro that matters the most. It is my hope that once 
the complexity of that forgotten territory between the micro and the macro 
is grasped at visceral level, the intellectual habit to privilege one or the 
other extreme will become easier to break. 

On the other hand, a solution to the micro-macro problem in terms of 
a multiplicity of social entities operating at intermediate levels of scale 
calls for a few words to clarify the meaning of the expression 'larger- 
scale'. Its usual meaning is geometric, as when when one says that a 
street is the longest one in a city, or that one nation-state occupies a larger 
area than another. But there is also a physical meaning of the expression 
that goes beyond geometry. In physics, for example, length, area and 
volume are classified as extensive properties, a category that also includes 
amount of energy and number of components. It is in this latter extensive 

sense, not the geometric one, that I use the expression 'larger-scale'. Two 
interpersonal networks, for example, will be compared in scale by the 
number of members they contain not by the extent of the geographical 
area they occupy, so that a network structuring a local community will be 
said to be larger than one linking geographically dispersed friends if it has 
more members, regardless of the fact that the latter may span the entire 
planet. Also, being larger in only one of the properties differentiating the 
social entities to be discussed here. There are many others properties 
(such as the density of the connections in a network, or the degree of 
centralization of authority in an organization) that are not extensive but 
intensive, and that are equally important. Finally, social entities will be 
characterized in this book not only by their properties but also by their 
capacities, that is, by what they are capable of doing when they interact 
with other social entities. 

To those readers who may be disappointed by the lack of cross-cultural 
comparisons, or the absence of detailed analyses of social mechanisms, or 
the poverty of the historical vignettes. I can only say that none of these 
worthy tasks can be really carried out within an impoverished ontological 
framework, When social scientists pretend to be able to perform these 
tasks without ontological foundations, they are typically using an 
implicit, and thereby uncritically accepted, ontology. There is simply no 
way out of this dilemma. Thus, while philosophers cannot, and should 
not, pretend to do the work of social scientists for them, they can greatly 
contribute to the job of ontological clarification. This is the task that this 
book attempts to perform. 

Manuel •eLanda 
New York, 2005 



Assemblages against Totalities 

The purpose of this chapter is to introduce the theory of assemblages. Bu 
this introduction is not meant as an end in itself, but as a means t< 
elucidate the proper ontological status of the entities that are invoked bj 
sociologists and other social scientists. Is there, for example, such a thinj 
as society as a whole? Is the commitment to assert the existence of suet 
an entity legitimate? And, is denying the reality of such an emit! 
equivalent to a commitment to the existence of only individual person: 
and their families?The answer to all these questions is a definitive no, bu 
several obstacles must be removed before justifying this negativ* 
response. Of all the obstacles standing in the way of an adequate soda 
ontology none is as entrenched as the organismic metaphor. In its leas< 
sophisticated form this stumbling-block involves making a superficial 
analogy between society and the human body, and to postulate that just 
as bodily organs work together for the organism as a whole, so the 
function of social institutions is to work in harmony for the benefit of 
society. As historians of social thought Howard Becker and Harry Barnes 
have noted, there are many variants of this centuries-old metaphor, some 
more sophisticated than others: 

The theory of the resemblance between classes, groups, and institu¬ 
tions in society and the organs of the individual is as old as social 
theory itself. We have already noted its presence in Hindu social 
thought, and have also called attention to the fact that Aristotle, in 
book IV of his Politics, sets forth this organismic analogy with precision 
and clarity. The same conception appears clearly in the writings o' 

Cicero, Livy, Seneca, and Paul. In the Middle Ages elaborate 
anthropomorphic analogies were drawn by John of Salisbury and 
Nicholas of Cues. In the early modern period, Hobbes and Rousseau 
contrasted the organism and the state, holding that the organism was 
the product of nature while the state was an artificial creation. In the 
late eighteenth and early nineteenth century fanciful notions of the 
social and political organism appeared with such writers as Hegel, 
Schelling, Krause, Ahrens, Schmitthenner, and Wait/. 1 

In the late nineteenth century the organismic metaphor achieved its 
first systematic development in the work of Herbert Spencer and reached 
its pinnacle of influence a few decades later in the work of Talcott 
Parsons, the most important figure of the functionalist school of 
sociology. After this, the use of the organism as a metaphor declined as 
sociologists rejected functionalism, some because of its emphasis on social 
integration and its disregard for conflict, others because of its focus on 
social structure at the expense of phenomenological experience. But a 
more sophisticated form of the basic metaphor still exerts considerable 
influence in most schools of sociology, and in this form it is much more 
difficult to eliminate. This version involves not an analogy but a general 
theory about the relations between parts and wholes, wholes that 
constitute a seamless totality or that display an organic unity. The basic 
concept in this theory is what we may call relations of interiority: the 
component parts are constituted by the very relations they have to other 
parts in the whole. A part detached from such a whole ceases to be what it 
is, since being this particular part is one of its constitutive properties. A 
whole in which the component parts are self-subsistent and their 
relations are external to each other does not possess an organic unity. 
As Hegel wrote: 'This is what constitutes the character of mechanism, 
namely, that whatever relation obtains between the things combined, 
this relation is extraneous to them that does not concern their nature at all, 
and even if it is accompanied by a semblance of unity it remains nothing 
more than composition, mixture, aggregation, and the like.' 2 

Thus, in this conception wholes possess an inextricable unity in which 
there is a strict reciprocal determination between parts. This version of 
organismic theory is much harder to eliminate because it is not just a 
matter of rejecting an old worn-out image and because its impact on 
sociology goes beyond functionalism. A good contemporary example is 
the work of the influential sociologist Anthony Giddens, who attempts to 



transcend the duality of agency and structure by arguing for their mutua 
constitution: agency is constituted by its involvement in practice which 
in turn, reproduces structure. Structure is conceived as consisting o 
behavioural procedures and routines, and of material and symbol! 
resources, neither one of which possesses a separate existence outside o 
their instantiation in actual practice. 3 In turn, the practices whicl 
instantiate rules and mobilize resources are conceived by Giddens as , 
continuous flow of action 'not composed of an aggregate or series o 
separate intentions, reasons, and motives'. 4 The end result of this is ; 
seamless whole in which agency and structure mutually constitute om 
another dialectically. 5 

Following Hegel, other defenders of this approach argue that withou 
relations of interiority a whole cannot have emergent properties 
becoming a mere aggregation of the properties of its components. I 
may be argued, however, that a whole may be both analysable int< 
separate parts and at the same time have irreducible properties, propertie 
that emerge from the interactions between parts. As the philosopher o 
science Mario Bunge remarks, the 'possibility of analysis does not entai 
reduction, and explanation of the mechanisms of emergence does no: 
explain emergence away'.* Allowing the possibility of complex interac 
tions between component parts is crucial to define mechanisms o 
emergence, but this possibility disappears if the parts are fused togethei 
into a seamless web. Thus, what needs to be challenged is the very idea o 
relations of interiority. We can distinguish, for example, the propertie! 
defining a given entity from its capacities to interact with other entities 
While its properties are given and may be denumerable as a closed list, it: 
capacities are not given - they may go unexercised if no entity suitable foi 
interaction is around - and form a potentially open list, since there is no 
way to tell in advance in what way a given entity may affect or be affected 
by innumerable other entities. In this other view, being part of a whole 
involves the exercise of a part's capacities but it is not a constitutive 
property of it. And given that an unexercised capacity does not affect 
what a component is, a part may be detached from the whole while 
preserving its identity. 

Today, the main theoretical alternative to organic totalities is what the 
philosopher Gilles Deleuze calls assemblages, wholes characterized by 
relations of exteriority. These relations imply, first of all, that a component 
part of an assemblage may be detached from it and plugged into a 
different assemblage in which its interactions are different. In other 

words the exteriority of relations implies a certain autonomy for the 
terms they relate, or as Deleuze puts it, it implies that 'a relation may 
change without the terms changing'/ Relations of exteriority also imply 
that the properties of the component parts can never explain the relations 
which constitute a whole, that is, 'relations do not have as their causes 
the properties of the [component partsj between which they are 
established -. .' 8 although they may be caused by the exercise of a 
component's capacities. In fact, the reason why the properties of a whole 
cannot be reduced to those of its parts is that they are the result not of an 
aggregation of the components' own properties but of the actual exercise 
of their capacities. These capacities do depend on a component's 
properties but cannot be reduced to them since they involve reference 
to the properties of other interacting entities. Relations of exteriority 
guarantee that assemblages may be taken apart while at the same time 
allowing that the interactions between parts may result in a true 

While those favouring the interiority of relations tend to use 
organisms as their prime example, Deleuze gravitates towards other 
kinds of biological illustrations, such as the symbiosis of plants and 
pollinating insects. In this case we have relations of exteriority between 
self-subsistent components - such as the wasp and the orchid - relations 
which may become obligatory in the course of coevolution. This 
illustrates another difference between assemblages and totalities. A 
seamless whole is inconceivable except as a synthesis of these very parts, 
that is, the linkages between its components form logically necessary 
relations which make the whole what it is. But in an assemblage these 
relations may be only contingently obligatory. While logically necessary 
relations may be investigated by thought alone, contingently obligatory 
ones involve a consideration of empirical questions, such as the 
coevolutionary history of two species. In addition to this Deleuze 
considers heterogeneity of components an important characteristic of 
assemblages. Thus, he would consider ecosystems as assemblages of 
thousands of different plant and animal species, but not the species 
themselves, since natural selection tends to homogenize their gene pools. 
In what follows I will not take heterogeneity as a constant property of 
assemblages but as a variable that may take different values. This will 
allow me to consider not only species but also biological organisms as 
assemblages, instead of having to introduce another category for them as 
does Deleuze. 9 Conceiving an organism as an assemblage implies that 


despite the tight integration between its component organs, the relations 
between them are not logically necessary but only contingently 
obligatory: a historical result of their close coevolution. In this way 
assemblage theory deprives organismic theories of their most cherished 

In addition to the exteriority of relations, the concept of assemblage is 
defined along two dimensions. One dimension or axis defines the variable 
roles which an assemblage's components may play, from a purely materia! 
role at one extreme of the axis, to a purely expressive role at the other 
extreme. These roles are variable and may occur in mixtures, that is, a 
given component may play a mixture of material and expressive roles by 
exercising different sets of capacities. The other dimension defines 
variable processes in which these components become involved and that 
either stabilize the identity of an assemblage, by increasing its degree of 
internal homogeneity or the degree of sharpness of its boundaries, or 
destabilize it. The former are referred to as processes of territorialization 
and the latter as processes of deterritorialization One and the same 
assemblage can have components working to stabilize its identity as well 
as components forcing it to change or even transforming it into a different 
assemblage. In fact, one and the same component may participate in both 
processes by exercising different sets of capacities. Let me give some 
simple social examples of these four variables. 

The components of social assemblages playing a material role vary 
widely, but at the very least involve a set of human bodies properly 
oriented (physically or psychologically) towards each other. The classic 
example of these assemblages of bodies is face-to-face conversations, but 
the interpersonal networks that structure communities, as well as the 
hierarchical organizations that govern cities or nation-states, can also 
serve as illustrations. Community networks and institutional organiza¬ 
tions are assemblages of bodies, but they also possess a variety of other 
material components, from food and physical labour, to simple tools and 
complex machines, to the buildings and neighbourhoods serving as their 
physical locales. Illustrating the components playing an expressive role 
needs some elaboration because in assemblage theory expressivity cannot 
be reduced to language and symbols. A main component of conversations 
is, of course, the content of the talk, but there are also many forms of 
bodily expression (posture, dress, facial gestures) that are not linguistic. In 
addition, there is what participants express about themselves not by what 
they say but by the way they say it, or even by their very choice of topic. 


These are nonlinguistic social expressions which matter from the point of 
view of a person's reputation (or the image he or she tries t* project in 
conversations) as much as what the person expresses linguistically. 
Similarly, an important component of an interpersonal network is the 
expressions of solidarity of its members, but these can be either linguistic 
(promises, vows) or behavioural, the solidarity expressed by shared 
sacrifice or mutual help even in the absence of words. Hierarchical 
organizations, in turn, depend on expressions of legitimacy, which may 
be embodied linguistically (in the form of beliefs about the sources of 
authority) or in the behaviour of their members, in the sense that the 
very act of obeying commands in public, in the absence of physical 
coercion, expresses acceptance of legitimate authority. 11 

The concept of territorialization must be first of all understood literally. 
Face-to-face conversations always occur in a particular place (a street- 
corner, a pub, a church), and once the participants have ratified one 
another a conversation acquires well-defined spatial boundaries. Simi¬ 
larly, many interpersonal networks define communities inhabiting spatial 
territories, whether ethnic neighbourhoods or small towns, with well- 
defined borders. Organizations, in turn, usually operate in particular 
buildings, and the jurisdiction of their legitimate authority usually 
coincides with the physical boundaries of those buildings. The exceptions 
arc governmental organizations, but in this case too their jurisdictional 
boundaries tend to be geographical: the borders of a town, a province or a 
whole country. So, in the first place, processes of territorialization are 
processes that define or sharpen the spatial boundaries of actual 
territories. Territorialization, on the other hand, also refers to non-spatial 
processes which increase the internal homogeneity of an assemblage, 
such as the sorting processes which exclude a certain category of people 
from membership of an organization, or the segregation processes which 
increase the ethnic or racial homogeneity of a neighbourhood. Any 
process which either destabilizes spatial boundaries or increases internal 
heterogeneity is considered deterritorializing. A good example is com¬ 
munication technology, ranging from writing and a reliable postal 
service, to telegraphs, telephones and computers, all of which blur the 
spatial boundaries of social entities by eliminating the need for co¬ 
presence: they enable conversations to take place at a distance, allow 
interpersonal networks to form via regular correspondence, phone calls 
or computer communications, and give organizations the means to 
operate in different countries at the same time. 






While the decomposition of an assemblage into its different parts, and 
the assignment of a material or expressive role to each component, 
exemplifies the analytic side of the approach, the concept of territor- 
ialization plays a synthetic role, since it is in part through the more or less 
permanent articulations produced by this process that a whole emerges 
from its parts and maintains its identity once it has emerged. But there is 
another synthetic process in assemblage theory that complements 
territorialization: the role played in the production and maintenance of 
identity by specialized expressive entities such as genes and words. 
Although Deleuze considers all entities, even nonbiological and nonsocial 
ones, as being capable of expression, he argues that the historical 
appearance of these specialized entities allowed a great complexification 
of the kinds of wholes that could be assembled in this planet. Let me 
elaborate this point starting with the idea that physical or chemical 
entities are capable of expression. When atoms interact with radiation 
their internal structure creates patterns in this radiation through the 
selective absorption of some of its wavelengths. In manmade photographs 
this pattern appears as a spatial arrangement of light and dark bands (a 
spectrograph) which is correlated in a unique way with the identity of the 
chemical species to which the atom belongs. In other words, the 
absorption pattern expresses the identity of the chemical species in the 
form of physical information which can be used by astrophysicists, for 
example, to identify the chemical elements present in a given celestial 
process. 12 

On the other hand, this expressivity is clearly not functional in any 
sense. That is, while the information patterns do have an objective 
existence, in the absence of astrophysicists (or other users of spectro¬ 
graphs) the patterns do not perform any function. These patterns may be 
compared to the fingerprints that are expressive of human organic 
identity, but that in the absence of a law-enforcement organization that 
collects them, stores them and retrieves them as part of a process of 
identification, perform no real biological function at all. But, Deleuze 
argues, there have been critical thresholds in the history of the planet 
when physical expressivity has become functional. The first threshold is 
the emergence of the genetic code, marking the point at which 
information patterns ceased to depend on the full three-dimensional 
structure of an entity (such as that of an atom) and became a separate 
one-dimensional structure, a long chain of nucleic acids. The second 
threshold is the emergence of language: while genetic linearity is still 

linked to spatial relations of contiguity, linguistic vocalizations display a 
temporal linearity that endows its information patterns with an even 
greater autonomy from their material carrier. 13 These two specialized 
lines of expression must be considered assemblages in their own right. 
Like all assemblages they exhibit a part-to-whole relation: genes are made 
up of linear sequences of nucleotides, and are the component parts of 
chromosomes; words are made of linear sequences of phonetic sounds or 
written letters, and are the component parts of sentences. Some of these 
component parts play a material role, a physical substratum for the 
information, and through elaborate mechanisms this information can be 
expressed as proteins, in the case of genetic materials, or as meanings, in 
the case of linguistic ones. 14 

In assemblage theory, these two specialized expressive media are 
viewed as the basis for a second synthetic process. While territorialization 
provides a first articulation of the components, the coding performed by 
genes or words supplies a second articulation, consolidating the effects of 
the first and further stabilizing the identity of assemblages. 15 Biological 
organisms are examples of assemblages synthesized through both 
territorialization and coding, but so are many social entities, such as 
hierarchical organizations. The coding process in the latter will vary 
depending on whether the source of legitimate authority in these 
hierarchies is traditional or rational-legal, as in modern bureaucracies. In 
the former the coding is performed by narratives establishing the sacred 
origins of authority, while in the latter it is effected by constitutions 
spelling out the rights and obligations associated with each formal role. It 
is tempting to see in the fact that both biological organisms and some of 
the most visible social institutions are doubly articulated, the source of 
the appeal of the organismic metaphor: the isomorphism of the processes 
giving rise to some biological and social entities would explain their 
resemblance. On the other hand, this real resemblance should not license 
the idea that 'society as a whole' is like an organism, since many social 
assemblages are not highly coded or highly territorialized. 

In fact, in both the biological and the social realms there are 
processes of decoding, yielding assemblages which do not conform to the 
organismic metaphor. In biology such decoding is illustrated by animal 
behaviour which has ceased to be rigidly programmed by genes to be 
learned from experience in a more flexible way. This decoding produces, 
for example, animal territories, the assemblages generated when 
animals have gone beyond the passive expression of information 


patterns (patterns of the fingerprint kind) actively to use a variety of 
means - from faeces and urine to song, colour and silhouette - as an 
expression of their identity as owners of a particular geographical area. 16 
A social example of the result of a process of decoding would be informal 
conversations between friends. As social assemblages, conversations do 
not have the same durability of either interpersonal networks or 
institutional organizations, and no one would feel tempted to compare 
them to organisms. But they do involve rules, such as those governing 
turn-taking. The more formal and rigid the rules, the more these social 
encounters may be said to be coded. But in some circumstances these 
rules may be weakened giving rise to assemblages in which the 
participants have more room to express their convictions and their 
own personal styles. 17 

Nevertheless, and despite the importance of genetic and linguistic 
components for the consolidation of the identity of biological and social 
assemblages, it is crucial not to conceptualize their links to other 
components as relations of interiority. In other words, the interactions 
of genes with the rest of a body's machinery should not be viewed as if 
they constituted the defining essence of that machinery. And similarly for 
the interactions of language with subjective experience or with social 
institutions. In an assemblage approach, genes and words are simply one 
more component entering into relations of exteriority with a variety of 
other material and expressive components, and the processes of coding 
and decoding based on these specialized lines of expression operate side by 
side with nongenetic and nonlinguistic processes of territorialization and 
deterritorialization. To emphasize this point in the chapters that follow, 1 
will always discuss language last and as a separate component. This will 
allow me to distinguish clearly those expressive components that are not 
linguistic but which are mistakenly treated as if they were symbolic, as 
well as to emphasize that language should be moved away from the core 
of the matter, a place that it has wrongly occupied for many decades now. 

There are two more questions that must be discussed to complete the 
characterization of the assemblage approach. The first regards the 
processes of assembly though which physical, biological and social 
entities come into being, processes that must be conceptualized as 
recurrent. This implies that assemblages always exist in populations, 
however small, the populations generated by the repeated occurrence 
of the same processes. As the assemblages making up these collectivities 
interact with one another, exercising a variety of capacities, these 


interactions endow the populations with some properties of their own, 
such as a certain rate of growth or certain average distributions of 
assemblage properties. The second question regards the possibility that 
within these collectivities larger assemblages may emerge of which the 
members of the population are the component parts. In other words, the 
interactions between members of a collectivity may lead to the formation 
of more or less permanent articulations between them yielding a macro¬ 
assemblage with properties and capacities of its own. Since the processes 
behind the formation of these enduring articulations are themselves 
recurrent, a population of larger assemblages will be created leading to 
the possibility of even larger ones emerging. 

The combination of recurrence of the same assembly processes at any 
one spatial scale, and the recurrence of the same kind of assembly 
processes (territorialization and coding) at successive scales, gives 
assemblage theory a unique way of approaching the problem of linking 
the micro- and macro-levels of social reality. The bulk of this book will be 
spent giving concrete examples of how we can bridge the level of 
individual persons and that of the largest social entities (such as territorial 
states) through an embedding of assemblages in a succession of micro- 
and macro-scales. But at this point it will prove useful to give a simple 
illustration. One advantage of the present approach is that it allows the 
replacement of vaguely defined general entities (like 'the market' or the 
state') with concrete assemblages. What would replace, for example, the 
market' in an assemblage approach? Markets should be viewed, first of 
all. as concrete organizations (that is, concrete market-places or bazaars) 
and this fact makes them assemblages made out of people and the 
material and expressive goods people exchange. 

In addition, as the economic historian Fernand Braudel argues, these 
organizations must be located in a concrete physical locale, such as a 
small town and its surrounding countryside, a locale which should also be 
considered a component of the assemblage. In these terms, the smallest 
economic assemblage has always been, as Braudel says. 

a complex consisting of a small market town, perhaps the site of a fair, 
with a cluster of dependent villages around it. Each village had to be 
close enough to the town for it to be possible to go to the market and 
back in a day. But the actual dimensions of the unit would equally 
depend on the available means of transport, the density of settlement 
and the fertility of the area in question. 18 




Roughly, prior to the emergence of steam-driven transport, the average 
size of these complexes varied between 160 and 170 square kilometres. In 
t e high Middle Ages, as European urbanization intensified, these local 
markets multiplied, generating a large population of similar assemblages 
Then, some of the market-places belonging to these population were 
assembled together into regional markets, larger assemblages with an 
average area of 1,500 to 1,700 square kilometres. Each such region 
typically exhibited a dominant city as its centre and a recognizable 
cultural identity, both of which are parts of the larger assemblage. Next 
came provincial markets, with dimensions about ten times as large as the 
regional markets they assembled, but a lesser degree of internal 
homogeneity. Finally, when several such provincial markets were 
stitched together, as they were in England in the eighteenth century, 
national markets emerged. 7 " 

This brief description yields a very clear picture of a series of differently 
scaled assemblages, some of which are component parts of others which 
m turn become parts of even larger ones. Although I left out the 
historical details behind the assembly of local market-places into regional 
markets, or those behind the creation of national markets, it is clear that 
in each case there was a process through which larger entities emerged 
from the assembly of smaller ones. As Braudel notes of national markets, 
they were a network of irregular weave, often constructed against all 
odds: against the over-powerful cities with their own policies, against the 
provinces which resisted centralization, against foreign intervention 
which breached frontiers, not to mention the divergent interests of 
production and exchanged The situation is, indeed, even more complex 
because I am leaving out long-distance trade and the international 
markets to which this type of trade gave rise. Bu. even this simplified 

picture is already infinitely better than the reified generality of 'the 

Let me summarize the main features of assemblage theory. First of all 
unlike wholes in which parts are linked by relations of interiority (that is' 
relations which constitute the very identity of the parts) assemblages are 
made up of parts which are self-subsistent and articulated by relations of 
exteriority, so that a part may be detached and made a component of 
another assemblage. Assemblages are characterized along two dimen¬ 
sions: along the first dimension are specified the variable roles which 
component parts may play, from a purely material role to a purely 
expressive one, as well as mixtures of the two. A second dimension 


characterizes processes in which these components are involved: 
processes which stabilize or destabilize the identity of the assemblage 
(territorialization and deterritorialization). In the version of assemblage 
theory to be used in this book, a third dimension will be added: an extra 
axis defining processes in which specialized expressive media intervene, 
processes which consolidate and rigidify the identity of the assemblage or, 
on the contrary, allow the assemblage a certain latitude for more flexible 
operation while benefiting from genetic or linguistic resources (processes 
of coding and decoding). 21 All of these processes are recurrent, and their 
variable repetition synthesizes entire populations of assemblages. Within 
these populations other synthetic processes, which may also be 
characterized as territorializations or codings but which typically involve 
entirely different mechanisms, generate larger-scale assemblages of 
which some of the members of the original population become 
component parts. 

To conclude this chapter I would like to add some detail to the 
description of the synthetic aspects of assemblage theory. In particular, to 
speak of processes of territorialization and coding which may be 
instantiated by a variety of mechanisms implies that we have an 
adequate notion of what a mechanism is. In the case of inorganic and 
organic assemblages these mechanisms are largely causal, but they do not 
necessarily involve linear causality, so the first task will be to expand the 
notion of causality to include nonlinear mechanisms. Social assemblages, 
on the other hand, contain mechanisms which, in addition to causal 
interactions, involve reasons and motives. So the second task will be to 
show what role these subjective components play in the explanation of 
the working of social assemblages. The first task is crucial because the 
shortcomings of linear causality have often been used to justify the belief 
in inextricable organic unities. In other words, the postulation of a world 
as a seamless web of reciprocal action, or as an integrated totality of 
functional interdependencies, or as a block of unlimited universal 
interconnections, has traditionally been made in opposition to linear 
causality as the glue holding together a mechanical world. Hence if 
assemblages are to replace totalities the complex mechanisms behind the 
synthesis of emergent properties must be properly elucidated. 

In addition to supplying an excuse for the postulation of a block 
universe, the formula for linear causality, 'Same cause, same effect, 
always', has had damaging effects on the very conception of the relations 
between causes and effects. In particular, the resemblance of that formula 


With the one for logical implication ('If c, then E necessarily') has misled! 
man philosophers into thinking that the relation between a cause and its i 
effect is basically that the occurrence of the former implies that of the 
ter. But if causality is to provide the basis for objective syntheses causal 
relations must be characterized as productive, that is. as a relation in which 

impliLT 22 Th CaUSC t > 7 d r an ° ther CVem (thC effect) ' not J ust 
P ; e events wh| ch are productively connected by causality 

causal", 51 >° r at ° miStiC CVemS SUCh 35 mechanical collisions. But 
usality may also connect complex entities, such as the component parts 

hat make up a whole. In this case, while the entity itself cannot act as a 

a cause sfnce" V ^ “ eVem ' * ^ “ S defining pr °P erties can 

ause. since changes, even simple quantitative ones, are events For the 

same reason, actions performed by a complex entity can also be causes 

Linear causality is typically defined in terms of atomistic events but 

o ce we depart from these we must consider the role that the internal 

- “ ° 311 e , mity P ,a * in the " a Y it is affected by an external 

ettem J n organization may, for example, determine that an 

effect a h In 86 ‘ menSity Wi " 3 lo -intensity effect (or no 

effect at all) and vice versa, that small causes may have large effects 

ah a w-1 3565 ° n0nhnear causa lity, defined by thresholds below or 
above which external causes fail to produce an effect, that is. thresholds 

JseTTlT 8 the . CapaC ‘ tieS ° f an entit y to be causally affected. In some 
ca es, this capacity to be affected may gain the upper hand to the point 

at external causes become mere triggers or catalysts for an effect As 
g e puts it in this case 'extrinsic causes are efficient solely to the 
tent to which they take a grip on the proper nature and inner processes 

differ'nf S ‘T ***** Vi °‘ ateS Since il im P Iies that 

erent causes can lead to one and the same effect - as when a switch 

om one internal state to another is triggered by different stimuli - and 
hat one and the same cause may produce very different effects 
dr pending on the part of the whole it acts upon - as when hormones 
stimulate growth when applied to the tips of a plant but inhibit it when 
applied to its roots. It is important to emphasize, however, that to refer 
o inner processes (or to an internal organization) does not imply that 
onhnear or catalytic interactions are examples of relations of interiority 
inner processes are simply interactions between the component parts of 
an en,„ v and do imply thal lhese parti a „ 

ese two departures from linearity violate the first part of the formula 
( same cause, same effect'), but the second part ('always ) may also be 



challenged. Violating this second part, the part involving strict necessity, 
resulis in statistical causality, a form of causality that becomes important 
the moment we start to consider not single entities but large populations 
of such entities. Thus, when one says that, in a given population of 
smokers, 'Smoking cigarettes causes cancer', the claim cannot be that one 
repeated event (smoking) produces the same event (the onset of cancer) 
in every single case. The genetic predispositions of the members of the 
population must also be taken into account, and this implies that the 
cause will produce its effect only in a high percentage of cases. 
Furthermore, statistical causality does not depend on the existence of 
complex internal processes in the members of a population. It may also 
obtain without such internal organization given that, outside of 
laboratory conditions, no series of events ever occurs in complete 
isolation from other series which may interfere with it. Thus, even if we 
had a population of genetically identical humans, smoking would still not 
always lead to the onset of cancer, since other activities (physical exercise, 
for example) may play a part in counteracting its effects. The most that 
one can say about external causes in a population is that they increase the 
probability of the occurrence of a given effect. 25 

It is clear that assemblage theory, in which assemblages can be 
component parts of other assemblages (leading to the internal organiza¬ 
tion behind nonlinear and catalytic causality), and in which assemblages 
are always the product of recurrent processes yielding populations 
(involving statistical causality), can accommodate these complex forms of 
causal productivity. And in doing so it takes away the temptation to use 
seamless-web imagery. For example, the idea that there are reciprocal 
forms of determination between parts can be accommodated via 
nonlinear mechanisms involving feedback (such as the negative feedback 
characterizing thermostats), mechanisms that do not imply a fusion 
between the parts of a whole. The chance encounters between 
independent series of events at the source of statistical causality can also 
contribute to eliminate totalities and the block universe they imply. As 
Bunge puts it: 

A further test of the falsity of the doctrine of the block universe is the 
existence of chance (that is, statistically determined) phenomena; 
most of them arise from the comparative independence of different 
entities, that is, out of their comparative reciprocal contingency or 
irrelevancy. The existence of mutually independent lines of evolution 



is in turn ensured by the attenuation of physical interactions with 

distance, as well as their finite speed of propagation - the most 

effective looseners of the tightness of the block universe. 2 * 

The two roles that components play in an assemblage, material and 
expressive, are related to these different forms of causality While 
material components include the entire repertoire of causal interactions, 
expressive ones typically involve catalysis. The odours, sounds or colours 
that territorial animals use as expressions of their identity, for example, 
act only as triggers for behavioural responses in both rivals and potential 
mates, both of which must possess complex nervous systems to be 
capable of being affected this way. This is also true of genes, many of 
which code for enzymes that are highly effective and specific catalysts, 
although genes also code for proteins which play a material role, such as 
being building-blocks for cellular membranes. Language, on the other 
hand, typically plays a catalytic role which assumes that both speakers 
and listeners have complex internal organizations. This internal order 
however, is only partially explained by material causes (such as 
possessing a nervous system) and implies more elaborate mechanisms. 
In particular, the capacity of human beings to be affected by linguistic 
triggers (as well as by nonlinguistic expressions of solidarity, legitimacy or 
prestige) demands explanations in which reasons for acting are involved 
and, in some cases, by explanations involving motives. Roughly, while 
reasons may be exemplified by traditional values or personal emotions, 
motives are a special kind of reason involving explicit choices and goals 27 
As the sociologist Max Weber argued long ago, causes, reasons and 
motives are typically combined in the interpretation of social action, that 
is, action oriented towards the behaviour of others. As he writes: 'A 
correct causal interpretation of a concrete course of action is arrived at 
when the overt action and the motives have been correctly apprehended 
and at the same time their relation has become meaningfully compre¬ 
hensible.' 2 * The fact that Weber speaks of 'causal interpretations' is 
conveniently ignored by most students of his method of understanding 
(or Verstehen). This method by no means licenses the conclusion that all 
social action may be read like a text, or that all social behaviour can be 
treated as an enacted document. 2 ’ The source of this mistaken assessment 
of Weber s method is a confusion of two different meanings of the word 
meaning : signification and significance, one referring to semantic content, 
the other to importance or relevance. That Weber had significance and 



not signification in mind when he wrote about 'meaningfully compre¬ 
hensible' social action is clear from the fact that he thought his method 
worked best when applied to cases involving matching means to an end, 
that is, social action involving choices and goals. 30 Understanding or 
making sense of such activities typically involves assessing the adequacy 
of the way in which a goal is pursued, or a problem solved, or the 
relevance or importance of a given step in the sequence. Some of these 
vvill be assessments of causal relevance when the sequence of actions 
involves interacting with material objects, as in the activities of black¬ 
smiths, carpenters or cooks. But even when it is not a matter of 
interacting with the material world, judgements about goal-oriented 
linguistic performance will typically be about the adequacy of a line of 
argument or the relevance of a piece of information, and not about 
semantics. Means-to-ends matching is an example of social action that 
demands motives as part of its explanation. 

What about the case of social action involving reasons? Some 
examples of this type of social action may not involve semantic 
interpretation at all. These are the cases in which the weight of tradition 
or the intensity of the feelings may be such that the social activities 
involved may lie 'very close to the borderline of what can justifiably be 
called meaningfully oriented action, and indeed often on the other 
side', 31 (The other side being social action explained in purely causal 
terms, as in reactions triggered by habitual or affective stimuli.) But there 
are other cases of explanation by reasons that do not reduce to causal 
ones and do not involve any deliberate choices by social actors. In these 
cases, making sense of social behaviour involves giving reasons such as 
the belief in the existence of a legitimate order, or the desire to live up to 
the expectations associated with that order. Beliefs and desires may be 
treated as attitudes towards the meaning of declarative sentences (that is, 
towards propositions), and to this extent they do involve reference to 
semantics. Propositional attitudes are also involved in social action 
explained by motives, of course, such as the belief in the causal adequacy 
of some means or the desirability of the goals. But in the case of 
traditional reasons for action, causal adequacy may not be a motivating 
factor, and the desirability of a course of action may not depend on 
specific goals. 32 It is only in this case that the relations between the 
propositions themselves, such as the relations between the propositions 
that make up a religious doctrine, become crucial to make sense of social 
activities. And yet even this case will demand a mixture of semantic 



interpretation of the sacred texts involved and of assessments of the 1 
relative importance of different portions of these texts for the explanation ] 
of concrete courses of action. 

Weber's method gives us a way to approach the question of ] 
mechanisms in social assemblages: mechanisms which will always j 
involve complex mixtures of causes, reasons and motives. 33 Not | 
acknowledging the hybrid nature of social mechanisms can be a source ] 
of misunderstanding and mystification in social science. For example, I 
social activities in which means are successfully matched to ends are | 
traditionally labelled 'rational'. But this label obscures the fact that these | 
activities involve problem-solving skills of different kinds (not a single j 
mental faculty like 'rationality') and that explaining the successful 
solution of practical problems will involve consideration of relevant J 
causal events, such as physical interactions with the means to achieve a • 

goal, not just calculations in an actor's head. Similarly, when giving j 

traditional routines as explanations one may reduce these to ritual and | 

ceremony (and label these 'irrational'), but this obscures the fact that j, 

many inherited routines are in fact problem-solving procedures which ' 

have been slowly refined by successive generations. These practical ; 

routines may be overlaid by ritual symbolism, while at the same time 
being capable of leading to successful causal interactions with material 
entities, such as domesticated plants and soil. 

In addition to preserving the objective and subjective components, 
social mechanisms must include the full variety of causal interactions, 
that is, they must take into account that the thresholds characterizing 
nonlinear causality may vary from one actor to another (so that the same 
external cause may affect one but not the other) and that causal 
regularities in the behaviour of individual actors are, as Weber himself 
argued, only probabilistic. 34 Statistical causality is even more important 
when we consider populations of actors. Thus, in the case of explanation 
by motives, we may acknowledge that individual actors are capable of 
making intentional choices, and that in some cases such intentional 
action leads to the creation of social institutions (such as the written 
constitutions of some modern nation-states), while at the same time insist 
that the synthesis of larger social assemblages is many times achieved as 
the collective unintended consequence of intentional action, that is, as a kind 
of statistical result. In the case of explanations by reasons, on the other 
hand, the collective aspect may already be taken into account if the 
beliefs and desires involved are the effect of socialization by families or 


schools. But this socialization must, in addition, be conceived in 
probabilistic terms. Much as the effects of genes on the bodily 
characteristics of plants and animals are a matter of probabilities (not 
linear causal determinism) and that, therefore, in describing populations 
w e are interested in the statistical distribution of the variation in these 
bodily properties, so the effects of socialization should always be pictured 
as variable and the proper object of study should be how this variation is 
distributed in a given population. 

This concludes the introduction of assemblage theory. The next 
chapter will add the only component which I left out here (the 
topological diagram of an assemblage) after which the ontological status 
of assemblages will be properly elucidated. It will also expand the 
discussion of the part-to-whole relation that figures so prominently in the 
distinction between assemblages and totalities, and show in more detail 
how assemblage theory can help to frame the problem of the relation¬ 
ships between the micro- and the macro-levels of social phenomena. 
Once the problem has been correctly posed the other chapters will 
attempt to flesh out a solution. 


Assemblages against Essences 

Essentialism is the main reason offered by many social scientists to justify 
their rejection of realism. Postulating social entities with an enduring and 
mind-independent identity, these critics would argue, implies the 
existence of essences defining that identity. But what exactly are these 
essences supposed to be? While very few realists today would feel 
ontologically committed to assert the existence of eternal archetypes, 
there are subtler forms of essentialism in which essences are introduced 
when taxonomists reify the general categories produced by their 
classifications. It is therefore important to begin this chapter by explaining 
how assemblage theory does not presuppose the existence of reified 

Taxonomic essentialism, as opposed to its Platonic variety, may be 
traced back to the work of the great philosopher Aristotle, who created a 
method for the classification of entities into a three-level hierarchy: the 
genus, the species and the individual. For example, if the genus in 
question is 'animal', the method demands that we find specific differences 
which divide this genus into lower classes: for example, 'two-footed' and 
'four-footed' animals. This new level, in turn, can be divided into even 
lower classes by differences of differences. But here one must be careful, 
since as Aristotle says, 'it is not proper to say that an animal which has the 
support of feet, one sort we find with wings and another without them, if 
one is to express oneself correctly ... But it is correct to say so if one kind 
has cloven, and another has feet that are not cloven; for these are 
differences of foot .. .'* This method, when properly followed, leads us to 
the point where we cannot find any further differences and reach the 


assemblages against essences 

, , of a species: human or horse. These species may be further divided, 

‘7 u se since we can divide humans into those which are black or 
° f “ musical or not musical, just or unjust, but these are not necessary 
vvhl ' h ,t mere accidental combinations defining individuals with 
"hu,“ f,?I level o, species, o, „ ,he leve, o, who, 
2L philosophers call 'natural kinds-, .ha. we find ,he essence or ve„ 

na ',“„ ,e e™lu n hon",y theory, of course, rhis line of argument would be 
resected The properties differentiating one animal species from another 
shck to Aristotle's example, would be considered every bit a 
contingent as those marking the differences between organisms. The 
nerdes of species are the result of evolutionary processes that just as 
they occurred could have not occurred. The enduring identity of a given 
species is accounted for in terms of the different forms of natural selection 
(predators parasites, climate) that steer the accumulation of genetic 
materials in the direction of greater adaptability, as well as the process 
through which a reproductive community becomes separated into two 
progressively divergent communities until they cannot mate with on 
ZZl, wiile the 8 !,rst process yields the differentiating P= e cd 
species the second one. called 'reproductive isolation , makes those 
properties more or less durable by Cosing its gene pool to external genet c 
nows This isolation need not result in perfectly impermeable barney 
Many plant species, for example, maintain their ability to exchange gen 
with other plant species, so their identity is fuzzy in the long-run But 
even the defining boundaries of fully reproductively isolated animals h e 
ourselves may be breached through the use of biotechnology, or 
example, or through the action of retroviruses, a fact that confirms 

mntineent nature of the boundaries. 

addition io sharing the coniingency of ihri, anduring P™P™' 
organisms and species are also alike in iha, b„,h are born n dun 
reproductive isolation marks the threshold ol spec,anon, that ,s. the 
historical birth ol a new species, and extinction del,nek „s e,ua 
historical death. What this implies is that a biologica species is 
individual entity, as unique and singular as the 0 ^ ani ^^at com^se^ 
but larger in spatiotemporal scale. In other words, individual organisms 
are the component parts of a larger individual whole, not the particul r 
members of a general category or natural kind. 4 The same point app ies 
any other natural kind. For example, chemical species, as classified n 
periodic table of the elements, may be reified by a commitment to 



existence of hydrogen, oxygen or carbon in general. But it is possible to 
acknowledge the objectivity of the table while refusing to reify its natural 
kinds. Atoms of a given species would be considered individual entities 
produced by recurrent processes {processes of nucleosynthesis) taking 
place within individual stars. Even though, unlike organisms, these atoms 
display much less variation, the fact that they were born in a concrete 
process gives each of them a history. This implies that there is no need to 
be ontologically committed to the existence of 'hydrogen in general' but 
only to the objective reality of large populations of hydrogen atoms. 

The lesson from these two examples is that taxonomic essentialism 
relies on a very specific approach to yield its reified generalities: it starts 
with finished products (different chemical or biological species), discovers 
through logical analysis the enduring properties that characterize those 
products, and then makes these sets of properties into a defining essence 
(or a set of necessary and sufficient conditions to belong to a natural 
kind). To avoid reification we must instead focus on the historical 
processes that produce those products, with the term 'historical' referring 
to cosmological and evolutionary history in addition to human history. 
Assemblage theory, as outlined in the previous chapter, avoids taxonomic 
essentialism through this manoeuvre. The identity of any assemblage at 
any level of scale is always the product of a process (territorialization and, 
in some cases, coding) and it is always precarious, since other processes 
(deterritorialization and decoding) can destabilize it. For this reason, the 
ontological status of assemblages, large or small, is always that of unique, 
singular individuals. In other words, unlike taxonomic essentialism in 
which genus, species and individual are separate ontological categories, 
the ontology of assemblages is flat since it contains nothing but differently 
scaled individual singularities (or hacceities ). As far as social ontology is 
concerned, this implies that persons are not the only individual entities 
involved in social processes, but also individual communities, individual 
organizations, individual cities and individual nation-states. 

Natural kinds, on the other hand, are not the only source of 
essentialist myths. Aristotle begins his analysis at a level above that of 
natural kinds, with the genus 'animal', and via logical differentiation 
reaches the level of species ('horse', 'human'). The question is, if his 
species can be replaced by individual singularities, can the same be done 
to his genera? The answer is that the highest levels of biological 
classifications, that of kingdom (the level that includes plants and 
animals) or even phyla - including the phylum 'chordata' to which 

humans as vertebrate animals belong - need a different treatment. A 
phylum may be considered an abstract body-plan common to all 
vertebrates and, as such, it cannot be specified using metric notions such 
as lengths, area or volumes, since each realization of the body-plan will 
exhibit a completely different set of metric relations. Therefore only non- 
metric or topological notions, such as the overall connectivity of the 
different parts of the body, can be used to specify it. To put this 
differently, a body-plan defines a space of possibilities (the space of all 
possible vertebrate designs, for example) and this space has a topological 
structure. The notion of the structure of a space of possibilities is crucial in 
assemblage theory given that, unlike properties, the capacities of an 
assemblage are not given, that is, they are merely possible when not 
exercised. But the set of possible capacities of an assemblage is not 
amorphous, however open-ended it may be, since different assemblages 
exhibit different sets of capacities. 

The formal study of these possibility spaces is more advanced in 
physics and chemistry, where they are referred to as 'phase spaces . Their 
structure is given by topological invariants called 'attractors', as well as by 
the dimensions of the space, dimensions that represent the 'degrees of 
freedom', or relevant ways of changing, of concrete physical or chemical 
dynamical systems. 5 Classical physics, for example, discovered that the 
possibilities open to the evolution of many mechanical, optical and 
gravitational phenomena were highly constrained, favouring those 
outcomes that minimize the difference between potential and kinetic 
energy. In other words, the dynamics of a large variety of classical systems 
were attracted to a minimum point in the possibility space, an attractor 
defining their long-term tendencies. In the biological and social sciences, 
on the other hand, we do not yet have the appropriate formal tools to 
investigate the structure of their much more complex possibility spaces. 
But we may venture the hypothesis that they will also be defined as 
phase spaces with a much more complex distribution of topological 
invariants (attractors). We may refer to these topological invariants as 
universal singularities because they are singular or special topological 
features that are shared by many different systems. It is distributions of 
these universal singularities that would replace Aristotle's genera, while 
individual singularities replace his species. Moreover, the link from one to 
another would not be a process of logical differentiation, but one of 
historical differentiation, that is, a process involving the divergent evolution 
of all the different vertebrate species that realize the abstract body-plan. 


The taxonomic categories bridging the level of phyla to that of species 
would represent the successive points of divergence that historically 
differentiated the body-plan. 

In addition to the roles and processes described in the previous chapter 
assemblages are characterized by what Deleuze refers to as a diagram, a set 
of universal singularities that would be the equivalent of body-plan, or 
more precisely, that would structure the space of possibilities associated 
with the assemblage. 4 Thus, while persons, communities, organizations, 
cities and nation-states are all individual singularities, each of these 
entities would also be associated with a space of possibilities characterized 
by its dimensions, representing its degrees of freedom, and by a set of 
universal singularities. In other words, each of these social assemblages 
would possess its own diagram. In the previous chapter I showed how a 
reified generality like 'the market' could be replaced by a concrete 
historical entity such as a national market: an entity emerging from the 
unification of several provincial markets, each of which in turn is born 
from the stitching together of several regional markets, in turn the result 
of the historical union of many local market-places. Each of these 
differently scaled economic units must be regarded as an individual 
singularity bearing a relation of part-to-whole to the immediately larger 
one, much as organisms are related to species. What would be a social 
example of a diagram and its universal singularities? 

Max Weber introduced a classification for social entities in terms of 
what he called ideal types. In his analysis of hierarchical organizations, for 
example, he found that there are three different ways in which their 
authority may gain legitimacy: by reference to a sacred tradition or 
custom (as in organized religion); by complying with rational-legal 
procedures (as in bureaucracies); or by the sheer presence of a 
charismatic leader (as in small religious sects). 8 I will use this classification 
in another chapter and add more detail to the description of the three 
types. At this point, however, it is important to clarify their ontological 
status because the term 'ideal type' seems to suggest essences. But we can 
eliminate these essences by introducing the diagram of an authority 
structure. In this space of possibilities there would be three universal 
singularities defining 'extreme forms' that authority structures can take. 
The dimensions of the space, that is, the degrees of freedom of an 
authority structure, would include the degree to which an office or 
position in a hierarchy is clearly separated from the incumbent - rational- 
legal forms have the most separation, followed by the traditional and 


assemblages against essences 

, -inri the degree to which the activities of the 

^^lization are routinized - the charismatic form would have the least 
or8 pe of rou tinization, while the other two w.uld be highly routuuze . 
d Tsh rt. individual and universal singularities, each in its own wa 
„ !n , rhp assemblage approach to operate without essences. They also 
3 t Tne the proper use of analytical techniques in this approach. While in 
defl tic essentialism the role of analysis is purely logical, decomposing 
“rnuT into its component species by the successive discovery of 
Pessary differences, for example, in assemblage theory analysis must go 
h nd Cc nd involve causal intensions in reality, such as lesions 
nJde to aT organ within an organism, or the poisoning of enzymes 
' th' a cell followed by observations of the effect on the w o 
Saviour. These interventions are 

disentangled and Zluse the"entity under study may be composed of 
nans operating at different spatial scales and the correct scale mu t be 
located’ In short, analysis in assemblage theory is not conceptual bu 

causal concerned with the discovery of the actual mechanisms operating at 
causal, concer Qther hand , the topological structure defining 

a given spatial sea . d mechamsm . 

possibilities. 10 Causal and quasi-causal forms of analysis are use 
SSXS whirls hadfby the S^tb cemm^heady 

mechanisms through which actual minimization is 
separate case redundant. Both the productive causal rda 
the quasi-causal topological constraints were 

explanation of classical phenomena. This insight retains its validity whe 

approaching the more complex cases of biology and sociology 

Despite the complementarity of causal and quasi-causal forms o 
analysis in this book I will emphasize the former. Indeed, althou & h 
trv to give examples of the inner workings of concrete assemblage 
whenever possible, no attempt will be made to describe every causal 
mechanism in detail. On the other hand, it is important to 


these mechanisms should be properly conceptualized, particularly thosj 
mechanisms through which social wholes emerge from the interaction 
between their parts. The question of mechanisms of emergence has majo 
consequences for social theory because it impinges directly on th< 
problem of the linkages between the micro and the macro. This recalcitran 
problem has resisted solution for decades because it has been consistently 
badly posed. Assemblage theory can help to frame the problem correctly, 
thus clearing the way for its eventual solution - a solution that will 
involve giving the details of every mechanism involved. 

Posing the problem correctly involves, first of all, getting rid of the idea 
that social processes occur at only two levels, the micro- and the macro¬ 
levels, particularly when these levels are conceived in terms of reified 
generalities like 'the individual' and 'society as a whole'. The example of 
national markets given in the previous chapter shows that there may be 
more than two scales. If this is the case, then the terms 'micro' and 
'macro' should not be associated with two fixed levels of scale but used to 
denote the concrete parts and the resulting emergent whole at any given 
spatial scale. Thus, a given provincial market would be considered 'macro' j 
relative to its component regional markets, but 'micro' relative t« the 
national market. The same approach could be used to eliminate 'society 
as a whole' by bridging the smallest scale (that of individual persons) and 
the largest (that of territorial states) through a variety of intermediately 
scaled entities. Some contemporary sociologists have, in fact, proposed to 
frame the question of the micro-macro link in just these terms, breaking 
with a long tradition of privileging one of the two sides of the equation. 11 
Given that at each scale one must show that the properties of the whole 
emerge from the interactions between parts, this approach may be 
characterized as ontologically 'bottom-up'. But does such a bottom-up 
approach, coupled with the assumption that individual persons are the 
bottom-most level, commit us to the methodological individualism of 
microeconomics? No, and for several reasons. 

First of all, methodological individualists invoke reified generalities 
('the rational individual’) and use them in an atomistic way: individuals 
making rational decisions on their own. In assemblage theory persons 
always exist as part of populations within which they constantly interact 
with one another. But more importantly, while the identity of those 
persons is taken for granted in microeconomics, in assemblage theory it 
must be shown to emerge from the interaction between subpersonal 
components. Just what these components are I will specify in the next 



, , , r nnw it is enough to point out that they exist and that, if 
Ch3 d be they may be considered the smallest social scale. In addition 
° mWage theory departs from methodological individualism m that it 
assemblage t^e y P as an assemblage that may become 

C o nC Txified al peTsons become parts of larger assemblages: in conversa- 
C ° ml (and other social encounters) they project an image or persona, m 
U °,works they play informal roles; and in organizations they acquire 
T roles- aUd they may become identified with these roles and 
forma ' , , of tbe j r identity In other words, as larger 

Petii rbla g es a emergeTom the interactions of their component parts, the 
StXe parts may acquire new layers a s the emergent whole reacts 

baC Gr a anting for the "nte being that the emergence of objectivity canbe 
given an appropriate markets to 

move up'from'th^bottot-most level? The problem with that example is 


organizations and then function as component parts. None 
suggests a simple Russian-doll relation. onalnetvv#rks may 

organizations, in torn. ,end ,o form arge, assemi 8 
hierarchies ^ernrn^—- ^ hm as „ Pro 

Z rr ^iions ,o srainiizo 



it and perform specialized functions, such as lobbying, in the case of 1 
special interest organizations, or collective bargaining, in the case of ■ 
unions and other worker associations. That is, social movements are a 9 
hybrid of interpersonal networks and institutional organizations. And | 
similarly for government hierarchies, which at each jurisdictional scale I 
must form networks with nongovernmental organizations in order to be 
able to implement centrally decided policies. 

All of these larger assemblages exist as part of populations: populations 
of interpersonal networks, organizations, coalitions and government 
hierarchies. Some members of these populations carry on their interac¬ 
tions within physical locales, such as neighbourhoods, cities or territorial 
states, while others may take a more dispersed form interacting with each 
other at a distance thanks to communication and transportation 
technologies. The physical locales themselves, being spatial entities, do 
tend to relate to each other in a simple way: neighbourhoods are 
composed of many residential, commercial, industrial and governmental 
buildings; cities are composed of many neighbourhoods; and territorial 
states are composed of many cities, as well as of rural villages and 
unpopulated areas. But this apparent simplicity disappears when we add 
to these locales the recurring social activities taking place in them. Thus, a 
given city will include in its component parts not only neighbourhoods 
but the communities and organizations inhabiting those neighbourhoods. 

It will also include many interpersonal networks existing in dispersed 
farm, that is, networks not structuring well-defined, localized commu¬ 
nities, as well as organizations without a hierarchical structure (such as | 
market-places) and thus without a well-defined spatial jurisdiction or a 
homogenous internal composition. 

It is possible, however, to preserve the insight that a reified generality 
like "society as a whole' can be replaced by a multiscaled social reality, as 
long as the part-to-whole relation is correctly conceptualized to 
accommodate all this complexity. First of all, although a whole emerges 
from the interactions among its parts, once it comes into existence it can 
affect those parts. As the philosopher Roy Bhaskar has argued, emergent 
wholes are real because they are causal agents capable of acting back on 
the materials out of which they are formed'. 1 * In other words, to give a 
complete explanation of a social process taking place at a given scale, we 
need to elucidate not only micro-macro mechanisms, those behind the 
emergence of the whole, but also the macro—micro mechanisms through 
which a whole provides its component parts with constraints and resources, 

placing limitations on what they can do while enabling novel 
performances. 13 In the networks characterizing tightly knit communities, 
for example, a variety of resources become available to their members, 
from physical protection and help to emotional support and advice. But 
the same density of connections can also constrain members. News about 
broken promises, unpaid bets and other not-honoured commitments 
travels fast in those networks: a property that allows them to act as 
enforcement mechanisms for local norms. Similarly, many hierarchical 
organizations have access to large reservoirs of resources, which can be 
available to persons occupying certain formal positions in its authority 
structure. But the regulations defining the rights and obligations of these 
formal positions act as constraints on the behaviour of the incumbents. 
Because the capacities of a whole to constrain and enable may go 
unexercised, it would be more accurate to say that they afford their 
component parts opportunities and risks, such as the opportunity to use a 
resource (an opportunity that may be missed) or the risk of violating a 
limit (a risk that may never be taken). 

Do these conclusions still hold when we deal with assemblages that do 
not have a well-defined identity, that is that do not possess either clear 
boundaries or a homogenous composition, such as low-density, dispersed 
interpersonal networks, or organizations in which decision-making is not 
centralized? The answer is that they do, but there are some important 
differences. In particular, these more or less deterritorialized assemblages, 
to use the previously introduced terminology, can still provide their 
components with resources, although they have a diminished capacity to 
constrain them. In a dense network in which everybody knows 
everybody else and people interact in a variety of roles, the information 
that circulates tends to be well known to all participants. It follows that a 
novel piece of news will probably come not from one of its component 
members but from someone outside the network, that is, from someone 
connected to members of the network through a weak link. This is the 
basis of the famous argument about the strength of weak links. 14 Low- 
density networks, with more numerous weak links, are for this reason 
capable of providing their component members with novel information 
about fleeting opportunities. On the other hand, dispersed networks are 
less capable of supplying other resources (e.g. trust in a crisis) that define 
the strength of strong links. 15 They are also less capable of providing 
constraints, such as enforcement of local norms. The resulting low degree 
of solidarity, if not compensated for in other ways, implies that as a 


whole, dispersed communities are harder to mobilize politically and less 
likely to act as causal agents in their interactions with other communities. 

A similar point applies t o institutional organizations in which decision¬ 
making is not centralized, such as local market-places. Prior to the advent 
of national markets (as well as department stores, supermarkets, and so 
on) market-places supplied their component parts with resources: they 
provided rural inhabitants with the opportunity to sell their goods and 
the town's residents with the opportunity to purchase them. In addition, 
local markets were the places where 'townspeople met, made deals, 
quarreled, perhaps came to blows ... All news, political or otherwise, was 
passed on in the market'. 16 In other words, market-places were the place 
where people linked weakly to one another had an opportunity to pass 
novel information. They also provided constraints, in the sense that the 
prices at which goods were traded were typically determined imperson¬ 
ally by demand and supply; while the decisions to buy and sell were 
intentional, prices emerged as a collective, unintended consequence of 
those intentional actions and imposed themselves on the actors. 17 But 
prices are a weaker constraint than formal regulations, and in any case 
they only constrain those buyers and sellers who do not have economic 

In addition to the capacity of wholes t* enable and limit their parts 
there are the causal capacities they exercise when they interact with one 
another. Thus, as I said above, the communities structured by networks 
may interact among themselves to form a political coalition, and some 
organizations may interact as part of larger governmental hierarchies. 
These larger assemblages are emergent wholes in the sense just defined: 
being part of a political coalition provides a community with resources, 
like the legitimacy derived from numerousness and unity, but it also 
constrains it to struggle only for those goals that the whole coalition has 
agreed on pursuing; local regulatory agencies participating in the 
implementation of a nationwide policy are provided by the central 
government with financial resources, while at the same time being legally 
constrained to operate in a subordinate position. It may be objected, 
however, that these alliances and subordinations are not the effect of 
these larger assemblages, but of the activities of the people that compose 
them: the alliances are created by individual activists acting as 
representatives of their communities, and the authority of a government 
agency with national jurisdiction over another with local jurisdiction is 
always exercised by individual officials. But it is possible to accept that 



assemblages of people must interact by means of the activity of people 
and at the same time argue that these larger entities do have their own 
causal capacities. The device that allows such a compromise is the concept 

of redundant causality. 

In the explanation of a concrete social process it may not be 
immediately clear whether the causal actors are the micro-components 
or the macro-whole. The ambiguity can be eliminated if there are many 
equivalent explanations of the process in question at the micro-leve or 
example, if a coalition between communities which was in fact create y 
the negotiations between a specific group of activists could have been 
created by negotiations among other alternative activists. In other words 
we may be justified in explaining the emerging coalition as the result of 
the interaction between entire communities if an explanation of the 
micro-details is unnecessary because several such micro-causes would 
have led to a similar outcome. 1 * In the same way, a large organization 
may be said to be the relevant actor in the explanation of an 
interorganizaiional process if a subs,Option oi the people occupy, n e 
specific roles in its authority structure leaves the organizational pohcie 
and its daily routines intact. Such a substitution would, of course, have to 
respect specialties (managers replaced by other managers, accountants by 
accountants, engineers by engineers), but if the emergent properties and 
capacities of the organization remain roughly the same after such 
change then it would be redundant to explain the interorgamzat.onal 
outcome by reference to specific managers, accountants and engineers 
when reference to many other such specialists would have left the 

outcome approximately invariant. 

And the same point applies to larger assemblages. Cities inte 
causally with one another by competing for immigrants from rura 
regions, for natural resources such as water or agricultural land and 
economic investment. Large cities, for example, can cast a causa 
shadow' over their surroundings, inhibiting the formation of new towns 
within their sphere of influence by depriving them of people, resources or 
trade opportunities. But, of course, it is not the cues as physical entities 
that can interact this way, but cities as locales for the activities of 
inhabitants, including merchants, investors and migrants, as well as 
market-places and government organizations. So why not say 
the interactions between the performers of these activities that cause one 
urban centre to inhibit the growth of another? Because if we replaced the 
merchants by other merchants, the market-places, by other market- 


places and so on, a very similar inhibiting effect would be achieved. On 
the other hand, if such a replacement led to a very different outcome that 

would be evidence that the phenomenon in question must be explained 
by mechanisms operating at a smaller scale, and that it would involve not 
only causes but also reasons, and even motives. 

Thus social assemblages larger than individual persons have an 
objective existence because they can causally affect the people that are 
their component parts, limiting them and enabling them, and because 
they can causally affect other assemblages at their own scale. The fact that 
in order to exercise their causal capacities, internally as well as externally, 
these assemblages must use people as a medium of interaction does not 
compromise their ontological autonomy any more than the fact that 
people must use some of their bodily parts (their hands or their feet, for 
example) to interact with the material world compromises their own 
relative autonomy from their anatomical components. And a similar 
point applies at larger scales. When cities go to war. a recurring event in 
the age of city-states, they interact causally through their military 
organizations. Whether this interaction should be viewed as one between 
organizations or between urban centres is a question to be answered in 
terms of causal redundancy. If a war lasts so long, or is fought at such a 
large scale, that strategic decision-making at the organizational scale 
matters less than the exhaustion of urban resources (recruits, weapons, 
food supplies), then it would make sense to view the episode as one 
involving an interaction between urban centres, since a substitution of 
one set of military organizations for another would leave the outcome 
relatively unchanged. The military organizations could be seen as the 
medium through which warring cities (or territorial states) interact, much 
as individual officers in different branches of the military are the medium 
of interaction for the organizations themselves. 

There are three more adjustments that need to be made to the 
specification of assemblage theory to make it capable of adequately 
accounting for a multiscaled social reality. The first is a qualification of the 
very concept of emergence. I said above that one strategy to avoid 
reifying general categories was to focus on the process of production 
instead of the list of properties characterizing the finished product. This is, 

in fact, correct, but it runs the risk of placing too much emphasis on the 

historical birth of a particular assemblage, that is, on the processes behind 

the original emergence of its identity, at the expense of those processes 

which must maintain this identity between its birth and its death: no 


assemblages against essences 

animation would be able to keep its identity without the ongoing 
° r f er actions among its administrative staff and its employees: no city 
ltl uld keep its identity without ongoing exchanges among its pohtica , 

C om c and religious organizations: and no nation-state would survive 
^hom consent interactions between its capital city and its other urban 
centres In technical terminology this can be expressed by saying that 
territorializing processes are needed not only historically to produce he 
identity of assemblages at each spatial scale but also to maintain it in 
pnre of destabilizing processes of deterritonalization. 
second^qualtfication^is related to the first . argued in the prevtous 
chapter that assemblages are always produced by process 
recurrent and that this implies that they always exist in populations^ 
Given a population of assemblages at any one scale, other processes can 
then generate larger-scale assemblages using members of this popula to 

very few exceptions, organizations come into being in a w °r *« V 
nooulated by other organizations. Furthermore, while some parts mu 
p e ex st th V e whole, others may be generated by the maintenance 
processed of an already existing whole: while cities are composed o 
populations of interpersonal networks and organizations, it is simply not 
the case that these populations had to be there prior to the emergence of a 
dty"n fact most networks and organizations come into being as parts of 

relatively local battles of attrition to one based on battles of annihilation 
n wh h the entire resources of a nation were mobilized - » a good 
example of a process demanding a multiscaled explanation: .. involved 
? . takine place at the urban and national scale (the Frenc 

Revolution, which produced the first armies of motivated citizens instead 



of expensive mercenaries); causes and reasons at the organizational scale 
(the breaking-down of monolithic armies into autonomous divisions each 
with its own infantry, cavalry and artillery); and reasons and motives at 
the personal scale, since Napoleon's own strategic genius and charisma, 
amplified by his influential position in interpersonal networks, played a 
crucial catalytic role. 

Let me summarize this chapter’s argument so far. The ontological 
status of any assemblage, inorganic, organic or social, is that of a unique, 
singular, historically contingent, individual. Although the term 'indivi¬ 
dual' has come to refer to individual persons, in its ontological sense it 
cannot be limited to that scale of reality. Much as biological species are 
not general categories of which animal and plant organisms are members, 
but larger-scale individual entities of which organisms are component 
parts, so larger social assemblages should be given the ontological status 
of individual entities: individual networks and coalitions; individual 
organizations and governments; individual cities and nation-states. This 
ontological manteuvre allows us to assert that all these individual entities 
have an objective existence independently of our minds (or of our 
conceptions of them) without any commitment to essences or reified 
generalities. On the other hand, for the manoeuvre to work, the part-to- 
whole relation that replaces essences must be carefully elucidated. The 
autonomy of wholes relative to their parts is guaranteed by the fact that 
they can causally affect those parts in both a limiting and an enabling 
way, and by the fact that they can interact with each other in a way not 
reducible to their parts, that is, in such a way that an explanation of the 
interaction that includes the details of the component parts would be 
redundant. Finally, the ontological status of assemblages is two-sided; as 
actual entities all the differently scaled social assemblages are individual 
singularities, but the possibilities open to them at any given time are 
constrained by a distribution of universal singularities, the diagram of the 
assemblage, which is not actual but virtual. 

Given the crucial role that the part-to-whole relation plays in all this, 
to conclude this chapter I would like to clarify two further aspects of it. So 
far i have considered only questions of spatial scale, the whole being 
spatially larger to the extent that it is composed of many parts. But 
biological species, the example i used as a point of departure, also operate 
at longer temporal scales, that is, they endure much longer than their 
composing organisms and they change at a much slower rate. The first 
question is then: Is there a similar temporal aspect to the part-to-whole 

relation in social assemblages? Then there is the matter of special entities, 
in both the biological and social realms, that seem to operate in a scale-free 
wav These are the specialized lines of expression I mentioned in the first 
chapter, involving genetic and linguistic entities. On the one hand, genes 
and words, are more micro than the bodies and minds of persons. On the 
other, they can also affect macro-processes: genes define the human 
species as a whole, and words can define religions commanding belief by 
large portions of that species. The second question is: How do these 
special assemblages affect the part-to-whole relation? 

The first important temporal aspect of social assemblages is the relative 
duration of events capable of changing them. »oes it take longer to effect 
enduring and significant changes in organizations than in people, for 
example, or longer in cities than in organizations? Here we must first 
distinguish between changes brought about by causal interactions among 
social assemblages without any conscious intervention by persons (i.e. 
changes produced as collective unintended consequences of intentional 
action) from those which are the result of deliberate planning. The former 
case involves slow cumulative processes of the products of repeated 
interactions. For example, during the seventeenth and eighteenth 
centuries in Europe the authority structure of many organizations 
changed from a form based on traditional legitimacy to one based on 
rational-legal bureaucratic procedures. The change affected not only 
government bureaucracies, but also hospitals, schools and prisons. When 
studied in detail, however, no deliberate plan can be discerned, the 
change occurring through the slow replacement over two centuries of 
one set of daily routines by another. Although this replacement did 
involve decisions by individual persons — persons who may have simply 
imitated in one organization what was happening in another motivated 
by a desire for legitimacy - the details of these decisions are in most cases 
causally redundant to explain the outcome: an outcome better under¬ 
stood as the result of repeated interactions among the members of an 
organizational population. A similar point applies to changes in urban 
settlements: the interactions among towns, through trade and competi¬ 
tion for immigrants and investment, yield results over extended periods 
of time in which small initial advantages accumulate, or in which self- 
stimulating dynamics have time to amplify initial differences. 

Thus in changes not explainable by reference to strategic planning, 
relatively long time-scales can be expected for significant changes to take 
place. But what about the other case? »o planned changes at 


organizational or urban scales reduce to the characteristic duration of 
individual decision-making? Enduring and important changes in this 
other case always involve mobilization of internal resources, both material 
resources, such as energy or money, as well as expressive ones, such as 
solidarity or legitimacy. 1 believe it is safe to say that the larger the social 
entity targeted for change, the larger the amount of resources that must 
be mobilized. Given that resources are always scarce, this implies that 
spatial scale does have temporal consequences, since the necessary means 
may not be available instantaneously and may need to be accumulated 
over time, in addition, resource mobilization must be performed against a 
variety of sources of inertia at any given scale, from tradition and 
precedent to the entrenched interests of those that may be affected by a 
particular change. This implies that the larger the spatial scale of the 
change, the more extensive the alliances among the people involved have 
to be, and the more enduring their commitment to change has to be. Let 
me illustrate this with two examples at different spatial scales: resource 
mobilization performed within an organization to change the organiza¬ 
tion itself, and resource mobilizations performed in a hierarchy of 

organizations to effect change at the scale of neighbourhoods or entire 

The first case, interorganizational change, may be illustrated by the 
need f*r organizations to keep up with rapid technological developments. 
Given a correct assessment by people in authority of the opportunities 
and risks of new technologies, can an organization change fast enough to 
time internal changes to external pressures? Or more simply, can the resources 
available to an organization be mobilized at will? in large, complex 
organizations this may not be possible. Changes in the way an 
organization operates are bound to affect some departments more than 
others, or withdraw resources from one department to endow another, 
and this will generate internal resistance which must be overcome 
through negotiation. The possibilities of success in these negotiations, in 
turn, will depend on the extent to which the formal roles in an authority 
structure overlap with the informal roles of the interpersonal networks 
formed by employees. If a network property (such as the centrality or 
popularity of a node) fails to coincide with formal authority, the result 
may be conflict and stalemate in the mobilization of resources 20 This 
means that even in the case where the decisions to change have been 
made by people who can command obedience from subordinates, the 
very complexity of joint action implies delays in the implementation of 



the centrally decided plans, and thus, longer time-scales for organiza¬ 
tional change. 

The effect of time-lags produced by the need to negotiate and secure 
compliance with central decisions becomes more prominent at larger 
spatial scales, as in the case of changes at urban levels brought about by 
the policies of a national government. The implementation of policies 
decided upon by legislative, executive or judicial organizations typically 
involves the participation of many other organizations, such as bureau¬ 
cratic agencies. These agencies can exercise discretion when converting 
policy objectives into actual procedures, programmes and regulations. 
Thus it is necessary to obtain their commitment to a given policy's 
objectives, and this commitment will vary in different agencies from 
intense concern to complete indifference. This introduces delays in the 
implementation process, as the necessary negotiations take place. These 
delays, in turn, mean that agencies not originally involved have time to 
realize they have jurisdiction over portions of the programme, or to assess 
that the policy in question will impinge on their interests. If these other 
agencies get involved they complicate the implementation process by 
adding to the number of veto-points that must be cleared. Implementa¬ 
tion then becomes a process of continuous adjustment of the original 
objectives to a changing political reality, with each adjustment involving 
delays in the negotiation and securing of agreements. Historically, failures 
to meet the original objectives of a policy have often reflected 'the 
inability of the implementation machinery to move fast enough to 
capture the agreements while they lasted'. 21 

A. second temporal aspect of social assemblages is their relative 
endurance: a question fundamental in sociology, given that one could 
hardly use the term ‘institution' to refer to a social phenomenon which 
did not last longer than a human life. People are normally born in a world 
of previously existing institutions (both institutional norms and organiza¬ 
tions) and die leaving behind many of those same institutions. But 
beyond mere longevity, we would want to know whether the processes 
that constantly maintain the identity of social assemblages yield a 
characteristic life-span correlated with different spatial scales. In other 
words, is large spatial extension correlated with long temporal duration? 
The answer is that there is no simple correlation. Interpersonal networks 
vary in duration: dispersed friendship networks do not endure longer 
than the persons that compose them, but tightly knit networks of 
neighbours living in proximity do yield communities that survive the 




death of their parts. The durability of institutional organizations also 
varies: on the low side, restaurants have an average life-span of only a 
few years (a fact that gives them a reputation as the 'fruit-flies' of the 
organizational world) but some religious, governmental and even 
economic organizations can last for several centuries. Cities, in turn, 
while also having a range of durations, have instances that have endured 
for millennia, a nd most of them tend to outlive many of the organizations 
they house. Finally, although some territorial states, such as large 
empires, have demonstrated a resilience allowing them to endure at least 
as long as cities, nation-states are much too young to know just how 
enduring they can be. Thus, in some cases spatial and temporal scales do 
correlate, but not in others. On the other hand, most social assemblages 
larger than people do tend to outlive them on average even today when 
rates of infant mortality have decreased and average human life 
expectancies increased. 

In the case of dense interpersonal networks, part of the explanation for 
their relatively longer life-spans is that their continuity is maintained by 
the overlap of successive generations of neighbours. Similarly, in the case 
of hierarchical organizations, changes of personnel are never total, that is, 
there is always an overlap between staff familiar with the daily routines 
and new employees. But in addition to this temporal overlap there is 
transmission of semantic information across generations, about the 
traditions and customs of a particular community, or about the formal 
and informal rules defining positions of authority in a particular 
hierarchical organization. This transmission of linguistic materials helps 
maintain the identity of social assemblages across time much as the flow 
of genetic materials helps to preserve the identity of biological 
assemblages. As I said in the previous chapter, these specialized media 
of expression must themselves be considered assemblages, inhabiting the 
planet not as single general entities but as populations of concrete 
individual entities in part-to-whole relations: populations of individual 
sounds, words and sentences; populations of individual nucleotides, 
genes and chromosomes. 

On the other hand, these assemblages are special in two ways. In the 
first place, they are capable of variable replication, through a physical 
template mechanism in the case of genetic materials, and through 
enforced social obligation in the case of linguistic materials. Populations 
of replicators, when coupled to any filter or sorting device, are capable of 
guiding change over time, allowing the weight of the past to impinge on 

the present. When the sorting device biases this evolution towards 
adaptation, populations of replicators can act as a learning mechanism, a 
means to track changes in an environment through their own internal 
changes. In the second place, these specialized assemblages are capable of 
operating at multiple spatial scales simultaneously: genes are active within 
cells, govern the functioning of organs, influence the behaviour of entire 
organisms, and obstacles to their flow define the reproductive isolation of 
a species; language shapes the most intimate beliefs of persons, the public 
content of conversations, the oral traditions of small communities, and 
the written constitutions of large organizations and entire governments. 22 
Thanks to the flow of linguistic replicators, assemblages operating at 
dillerent spatial scales may also replicate, as when an organization opens 
a new branch in a different locality and sends part of its staff to transmit 
the daily routines defining its activities to the new employees. But the 
flow of linguistic replicators need not always be 'vertical' from one 
generation to another of the same community, or from one organization 
to a new branch. As with poorly reproductively isolated micro-organisms, 
this flow may be 'horizontal', introducing alien routines, procedures or 
rituals which alter, rather than preserve, the identity of social 

These characteristics make genetic and linguistic assemblages not 
ordinary assemblages. But however special, they should never be 
considered as any more than component parts entering into relations of 
exteriority with other component parts. When these relations are 
conceived as interiority relations, constitutive of the very identity of the 
related parts, both genes and words degenerate into essences. In the case 
of language this manoeuvre is embodied in the thesis of the linguisticality 
of experience, that is, the idea that an otherwise undifferentiated 
phenomenological field is cut up into discrete entities by the meanings 
of general terms. Since in many cases the meaning of general categories is 
highly stereotyped (particularly when they are categories applying to 
groups of people, as in gender or race categories) the thesis of the 
linguisticality of experience implies that perception is socially con¬ 
structed. 23 At the start of this chapter I argued that general categories do 
not refer to anything in the real world and that to believe they do (i.e. to 
reify them) leads directly to essentialism. Social constructivism is 
supposed to be an antidote to this, in the sense that by showing that 
general categories are mere stereotypes it blocks the move towards their 
reification. But by coupling the idea that perception is intrinsically 


linguistic with the ontological assumption that only the contents of 
experience really exist, this position leads directly to a form of social 
essentialism. In the following chapters, as I perform a detailed analysis of 
social assemblages at progressively larger spatial scales, these dangers 
must be kept in mind, particularly at the outset as I attempt to explain 
how individual persons emerge from the interaction of subpersonal 
components, only some of which will turn out to be linguistic. 



Persons and Networks 

Although persons are not the smallest analytical unit that social science 
can study - actions by persons such as individual economic transactions 
can be used as units of analysis - they are the smallest-scale social 
assemblage considered here. It is true that persons emerge from the 
interaction of subpersonal components, and that some of these 
components may justifiably be called the smallest social entities. Nothing 
very important depends on settling this question. All we need is a point of 
departure for a bottom-up ontological model, and the personal scale will 
provide a convenient one. On the other hand, it must be stated at the 
outset that the goal cannot be to settle all the philosophical questions 
regarding subjectivity or consciousness: questions that will probably 
continue to puzzle philosophers for a long time to come. All that is 
needed is a plausible model of the subject which meets the constraints of 
assemblage theory, that is, a model in which the subject emerges as 
relations of exteriority are established among the contents of experience. 
A good candidate for such a model, as Deleuze himself argued long ago, 
can be found in the philosophical school known as empiricism. 

The empiricist tradition is mostly remembered for its epistemological 
claims, in particular, the claim that all knowledge, including verbal 
knowledge, can ultimately be reduced to sense impressions. Or what 
amounts to the same thing, that sense experience is the foundation of all 
knowledge. But Deleuze discovered in the work of David Hume 
something much more interesting than such a dated foundational 
epistemology: a model of the genesis of subjectivity that can serve as an 
alternative to the dominant one based on the thesis of the linguisticality 



of experience. An empiricist model conceptualizes subjective experience 
first and foremost in terms of distinct and separable sense impressions. The 
ideas we derive from those impressions (ideas which may constitute the 
meanings of some words) are not related to them via social conventions 
but are direct replicas of those impressions, distinguishable from them 
exclusively by their lower intensity. 1 From the point of view of assemblage 
theory, it is crucial that each type of impression - not only visual, aural, 
olfactory and tactile but also the plurality of passions, from pride and 
humiliation to love and hatred - possess its own singular individuality, 
that is, that each of these impressions is, as Hume says, 'an original 
existence '. 2 This guarantees their heterogeneity and their irreducibility to 
one another. In addition, the singular status of impressions is what 
distinguishes empiricist from language-based models in which a particular 
impression is recognized as being the impression of something by 
mentally classifying it as belonging to a general category. 

On the other hand, some process must give these singular impressions 
and ideas a certain unity, even if that implies increasing their degree of 
uniformity and constancy. This process, as is well known, is the 
association of ideas. Can it be modelled via relations of exteriority? I 
argued before that the action of causes on their effects provides a good 
instant iation of relations of exteriority. For similar reasons, the action of 
formal operators on their arguments also constitutes a good example. In the 
case of subjectivity certain operators acting on ideas produce associative 
links between them and, in the process, provide subjective experience 
with its overall coherence. More specifically, the habitual grouping of 
ideas through relations of contiguity (in space or time), their habitual 
comparison through relations of resemblance, and the habitual pairing of 
causes and effects by their perceived constant conjunction, turns a loose 
collection of individual ideas into a whole with emergent properties. The 
associative relations established between ideas by these three operators 
meet the criteria of exteriority because they may change without the 
ideas themselves changing, and the properties of the ideas are not used to 
explain the operations that are applied to them , 3 

These three associative operators must be conceived as common to all 
humanity, being, according to Hume, 'original qualities of human 
nature '. 4 Speaking of a shared 'human nature", of course, should not 
be taken to imply any commitment to essentialism, since the human 
species is as much a contingent historical production as any human 
organism. Species-wide properties, being much more long-lasting than 



those of organisms or persons, can indeed seem to involve a 'fixed, 
necessary nature' when considering events at the organismic temporal 
scale, but this fixity and necessity are a kind of 'optical illusion' produced 
by the much slower rate of change of a species' properties, or by its high 
degree of reproductive isolation. On the other hand, a process which 
accounts for the emergence of a species-wide form of subjectivity leaves 
*ut many features that characterize individual persons belonging to 
individual cultures. Thus, while the habitual association of causes and 
effects allows any human subject to match means to ends (i.e. to solve 
practical problems), the choice of ends depends entirely on the passions: 
on the habitual pursuit of those ends associated with pleasurable or 
positively valued passions, and the habitual avoidance of those linked 
with painful or negatively valued ones . 3 The subject that emerges from 
this double process is a pragmatic subject whose behaviour must be 
explained both by giving reasons, such as traditional values, as well as by 
stating personal motives. We may summarize this model of the 
emergence of subjectivity using Deleuze's own words; 

... what transforms the mind into a subject and constitutes a subject in 
the mind are the principles of human nature. These principles are of 
two kinds; principles of association and principles of passions, which in 
some respects, we could present in the general form of the principle of 
utility. The subject is the entity which, under the influence of the 
principle of utility, pursues a goal or an intention; it organizes means 
in view of an end and, under the influence of the principles of 
association, establishes relations among ideas. Thus, the collection 
becomes a system. The collection of perceptions, when organized and 
bound, becomes a system . 6 

This systematic entity may be treated as an assemblage by distinguish¬ 
ing those components playing a material role from those playing an 
expressive role, and those processes that give it stability from those that 
destabilize it. The material role is performed by the bodily mechanisms 
behind the production of sense impressions, those underlying the body's 
dispositions towards the wide range of human passions and emotions, 
and those that realize neurologically the three associative operators. 
Although Hume himself refused to speculate on the nature of these 
mechanisms he did believe that the basic impressions emerge 'from the 
constitution of the body, from the animal spirits, or from the application 



of objects totheexternal organs '. 7 To these mechanisms we must add the 
energy or labour that, in the form of focused attention, is involved in the 
continuous production of associative links. The expressive role, on the 
other hand, is played both by linguistic and nonlinguistic components. 
The main example of the latter are the ideas derived from both sensual 
and passionate impressions. As I remarked before, the link between ideas 
and impressions is not representational, that is, not mediated by a 
convention or a code. Ideas directly express impressions. As Hume puts it, the 
'idea of red, which we form in the dark, and that impression, which 
strikes our eyes in sunshine, differ only in degree, not in nature '. 8 

The main territorializing process providing the assemblage with a 
stable identity is habitual repetition. Habit, for Hume, is a more powerful 
force sustaining the association of ideas than conscious reflection, and 
personal identity is stable only to the extent that habitual or routine 
associations are constantly maintained.’ It follows that any process which 
takes the subject back to the state it had prior to the creation of fixed 
associations between ideas (i.e. the state in which ideas are connected as 
in a delirium ) can destabilize personal identity. Examples of these 
deterritorializing processes are not hard to find. They include madness, 
high fever, intoxication, sensory deprivation and even deliberate 
interventions aimed at disrupting daily routine, as performed, for 
example, on prisoners in concentration camps. These, and other 
processes, can cause a loss, or at least a severe destabilization, of 
subjective identity . 10 

Personal identity, on the other hand, may be deterritorialized not only 
by loss of stability but also by augmentation of capacities; here we must 
go beyond Hume and add to habit or routine the effects of the acquisition 
of new skills. When a young child learns to swim or to ride a bicycle, for 
example, a new world suddenly opens up for experience, filled with new 
impressions and ideas. The new skill is deterritorializing to the extent that 
it allows the child to break with past routine by venturing away from 
home in a new vehicle, or inhabiting previously forbidden spaces like the 
ocean. New skills, in short, increase one's capacities to affect and be 
affected, or to put it differently, increase one's capacities to enter into 
novel assemblages, the assemblage that the human body forms with a 
bicycle, a piece of solid ground and a gravitational field, for example. Of 
course, the exercise of a new skill can soon become routine unless one 
continues to push the learning process in new directions. In addition, 
while rigid habits may be enough to associate linear causes and their 



constant effects, they are not enough to deal with nonlinear causes that 
demand more adaptive, flexible skills. 

Finally, there is the question of the role played by those expressive 
components that are linguistic. These must be introduced respecting the 
constraint against relations of inferiority, a constraint that, as I said 
before, rules out a neo-Kantian constitutive role for language. Moreover, 
it must be kept in mind that language came relatively late in the history of 
the evolution of the human species. As an intelligent species we spent 
millennia successfully coping with environmental challenges using 
accumulated knowledge about cause-and-effect relations. Hume himself 
argues that the ability to match means to ends (i.e. the capacity for causal 
reasoning), is not an exclusively human ability but may be observed in 
other animals which use it 'for their own preservation, and the 
propagation of their species '. 11 So to be compatible with assemblage 
theory, any given account of language must be capable of explaining its 
first emergence on the basis of a prior nonlinguistic form of intelligence. 
On the other hand, when language finally emerged it augmented those 
prior forms of intelligent behaviour through its much greater combinator¬ 
ial productivity. One difficulty with the associationist approach, a difficulty 
often pointed out by its critics, is the move from simple ideas to more 
complex ones. In Hume's account, for instance, the complex idea of an 
apple would be produced by combining simple ideas for a certain colour, 
shape, aroma, taste, and so on. But this combinatorial capacity pales 
when compared to that of language: given a dictionary with a finite 
number of words, a set of grammatical rules can produce an infinite 
number of well-formed sentences . 12 From the point of view of 
assemblage theory there is no problem in simply adding this combinator¬ 
ial productivity of language to that of associationism, as long as the theory 
of grammar that accounts for it (and several existing theories do) can also 
pass the evolutionary test, (i.e. that the formal operators it postulates be 
capable of emerging from a prior nonlinguistic form of subjectivity ). 13 

Assuming that we have a linguistic theory that meets all the 
requirements, the main effect of language at the personal scale is the 
shaping of beliefs. In the Humean account, the difference between belief 
and disbelief relative to a given idea is simply a question of intensity . 14 
Given that ideas are low-intensity replicas of impressions, believing in 
them simply brings them closer to those impressions, which is why, 
according to Hume, '[an] idea assented to feels different from a fictitious 
idea ’. 15 This notion of belief as feeling contrasts sharply with that of 




modern philosophers who stress the role of language. As I said in Chapter 
1, a belief may be considered an attitude towards a proposition, that is, 
towards the meaning of a sentence stating (truly or falsely) matters of 
fact. Given that declarative sentences are an important example of the 
combinatorial productivity of language, and that within assemblage 
theory this productivity is accepted as real, we must take seriously the 
definition of belief as a propositional attitude. On the other hand, this 
does not rule out the Humean notion since we may adopt such an 
attitude with different degrees of intensity, and in many cases it is the 
intensity of a given belief, more that the corresponding proposition, that 
drives social action. Thus, some people may be willing to die for a cause if 
they believe that martyrdom will guarantee eternal rewards, but this 
willingness to sacrifice themselves is more closely linked to the intensity 
of their devotion than to the specific semantic content of the belief (say, a 
certain number of virgins waiting in heaven), a content which could be 
altered without altering the behaviour. The importance of intensity over 
semantic content is clearer in other propositional attitudes, such as desires, 
which may take as objects purely Humean ideas (a desire for a specific 
taste or sound, or a particular visual experience) although they may also 
be directed towards propositions, as in the case of a desire for eternal 

This brief sketch of how subjectivity may be handled within 
assemblage theory can hardly be the last word, but it will be enough to 
provide us with a point of departure. The subject or person emerging from 
the assembly of subpersonal components (impressions, ideas, proposi¬ 
tional attitudes, habits, skills) has the right capacities to act pragmatically, 
(i.e. to match means to ends) as well as socially, to select ends for a 
variety of habitual or customary reasons that need not involve any 
conscious decision. On the other hand, given that the processes that 
produce assemblages are always iterative (i.e. that they always yield 
populations), we must immediately add those aspects of subjectivity that 
emerge from the interactions between persons. Some of these interac¬ 
tions may be viewed as taking place within assemblages, albeit ones with 
much shorter life-spans. These ephemeral assemblages may be referred to 
as 'social encounters', and of the many different types of social 
encounters we may single out a particularly relevant one: conversations 
between two or more persons. 

The most important research in this regard is without doubt 
represented by the work of the sociologist Erving Goffman who studied 

[ the way in which conversations add another layer of identity to persons: 
the public image or persona they project in their encounters with others, an 
image that has less to do with who they are than who they want to be. 
Goffman’s analysis of conversations lends itself to an assemblage 
approach first of all for its stress on relations of exteriority. He defines 
his subject matter as 

the class of events which occurs during co-presence and by virtue of 
co-presence. The ultimate behavioral materials are the glances, 
gestures, positionings, and verbal statements that people continuously 
feed into the situation, whether intended or not. These are the external 
signs of orientation and involvement - states of mind and body not 
ordinarily examined with respect to their social organization. 16 

In addition, Goffman's approach emphasizes the properties of 
conversations that cannot be reduced to their component parts, such as 
that of possessing a ritual equilibrium which may be threatened if 
involvement or attention are not properly allocated. A good example of 
threats to ritual stability are embarrassing events, such as linguistic errors 
(mispronunciation or misuse of words, lack of availability of a word when 
needed) or breaches of etiquette (making fun of a stutterer, calling a 
misstatement a lie), since these incidents divert attention from the 
conversation itself to the norms which the participants mutually enforce. 
When such threats occur it is the situation itself that becomes 
embarrassing: while the participant who committed the error may feel 
humiliated, particularly if the others do not allow him or her to save face, 
the other participants may also feel embarrassed about the incident itself, 
so that the entire situation suffers and must be repaired. The degree to 
which repair is necessary is directly linked to the intensity of the 
disruption. As Goffman writes, a humiliating event places all participants 
in 'a state of ritual disequilibrium or disgrace, and an attempt must be 
made to reestablish a satisfactory ritual state for them ... The imagery of 
equilibrium is apt here because the length and intensity of the corrective 
effort is nicely adapted to the persistence and intensity of the threat.' 17 

As an assemblage, a conversation possesses components performing 
both material and expressive roles. The main material component is co¬ 
presence: human bodies correctly assembled in space, dose enough to hear 
each other and physically oriented towards one another. Another 
material component is the attention and involvement needed to keep 





the conversation going, as well as the labour spent repairing ritual 
disequilibrium. While in routine conversations this labour may consist of 
simple habits, other occasions may demand the exercise of skills, such as 
tact (the capacity to prevent causing embarrassment to others) and poise 
(the capacity to maintain one's composure under potentially embarras¬ 
sing circumstances). 18 These are the minimal components playing a 
material role. But technological inventions (such as telephones or 
computer networks) may make strict physical co-presence unnecessary, 
leading to the loss of some material components (spatial proximity), but 
adding others, the technological devices themselves as well as the 
infrastructure needed to link many such devices. 

Although the flow of words making up the content of a conversation 
clearly plays an important expressive role, every participant in a 
conversation is also expressing claims to a certain public persona through 
every facial gesture, bodily posture, deployment of (or failure to deploy) 
poise and tact, choice of subject matter, and in many other ways. The 
expression of these identity claims must be done carefully, that is, 
performed in such a way that the image projected cannot be easily 
discredited by others. Any given conversation will present its participants 
with opportunities to express favourable information about themselves, 
as well as with risks to express unfavourable facts. Since this information 
becomes part of one's reputation, it will affect the identity claims one can 
afford to express in the next encounter. The variety of means through 
which claims to a public persona can be expressed constitute the main 
nonlinguistic expressive component of these assemblages. 

A conversation may be said to be territorialized by behavioural 
processes defining its borders in space and in time. The spatial boundaries 
of conversations are typically well defined, partly because of the physical 
requirement of co-presence, partly because the participants themselves 
ratify each other as legitimate interactors and exclude nearby persons 
from intruding into the conversation. As Goffman puts it, when 'the 
process of reciprocal ratification occurs, the persons so ratified enter into a 
state of talk - that is, they have declared themselves officially open to one 
another for purposes of spoken communication and guarantee together 
to maintain a flow of words'. 19 Conversations also have boundaries in 
time, defined by conventional ways of initiating and terminating an 
encounter, as well as a temporal order specifying the taking of turns 
during the encounter. 

Any event, or series of events, that destabilizes the conversation or 

blurs its boundaries may be considered deterritorializing. Embarrassment 
or humiliation are examples of the former: to the extent that claims to a 
publicly acceptable self circulate in conversations, any damage done to 
these public images is a potential threat to the integrity of the situation. 
Goffman discusses critical points in the intensity of humiliation, for 
example, after which regaining composure becomes impossible, embar¬ 
rassment is communicated to all participants, and the conversation 
collapses. 20 Beyond this, there are events which may transform a casual 
conversation into a heated discussion and, if the situation is not corrected, 
into a fist-fight. These events are also deterritorializing. Finally, a 
technological invention that allows a conversation to take place at a 
distance affects its identity not by changing it into some other form of 
social encounter but by blurring its spatial boundaries, forcing partici¬ 
pants to compensate for the lack of co-presence in a variety of other ways. 

The role performed by language in these assemblages is straightfor¬ 
ward, given that what is communicated in these exchanges are words and 
sentences. But, as I argued in a previous chapter, these linguistic entities 
have both signification and significance, and these two dimensions of 
meaning, one semantic the other pragmatic, should never be confused. 21 
One way in which the pragmatic dimension of language can be brought 
up is by thinking of the consequences of saying something. As Goffman 
argues, with the possible exception of activities deliberately intended to 
kill time, all other human activities have pragmatic consequences. In 
many cases these consequences are relatively well known in advance, 
due to their very high probability of occurrence, and are not therefore 
problematic. But other situations are both consequential and problematic. 
These he calls eventful or fateful. 22 The term applies, of course, to many 
types of social encounters, some of which may have a minimal linguistic 
component, such as the encounter of enemy armies on a battlefield. But 
it may also be used to distinguish conversations in which routine and 
relatively insignificant words are exchanged from those in which a 
subject matter of great importance to the participants is being discussed, 
and in which the outcome of the discussion is not easy to predict in 
advance. From the point of view of the identity claims one can make in 
social encounters, eventfulness changes the distribution of opportunities 
and risks. In particular, only eventful situations allow participants the 
expressive possibility of displaying character (courage, integrity, sports¬ 
manship). This is a significant opportunity because eventful encounters 
are relatively rare and, if the opportunity is not missed, participants can 


enhance their reputations in a long-lasting manner, since claims to strong 
character can only be challenged by the occurrence of another 
problematic event, not by the other participants. 23 

When conversations (and other social encounters) are repeated with 
the same participants, or with overlapping sets of participants, longer- 
lasting social entities tend to emerge: interpersonal networks. From the 
assemblage point of view, interpersonal networks are perhaps the social 
entities that are the easiest to handle, given that in network theory the 
emphasis is always on relations of exteriority. That is, it is the pattern of 
recurring links, as well as the properties of those links, which forms the 
subject of study, not the attributes of the persons occupying positions in a 
network. These attributes (such as gender or race) are clearly very) 
important in the study of human interactions, but some of the emergent I 
properties of networks tend to remain the same despite changes in those] 
attributes. This implies that the properties of the links cannot be inferred 1 
from the properties of the persons linked. The properties of links include j 
their strength, that is, the frequency of interaction among the persons] 
occupying given positions, as well as the emotional content of the j 
relation; their presence or absence, the absences indicating the existence of 
borders separating one network from another, or one clique from another; 
within a given network; and their reciprocity, that is, the symmetry or 1 
asymmetry of the obligations entailed by the link. 24 

In addition, the overall network has properties of its own, one of the 
most important of which is density, a measure of the intensity of 
connectivity among indirect links. 25 Roughly, if the friends of my friends 
(that is, my indirect links) know the friends of your friends, and they 
know the friends of everybody else's friends in a given community, the 
network has a high density. As 1 argued in the previous chapter, in a 
high-density network information about transgressions of local norms 
becomes known to all members of a community, which implies that the 
network itself has the capacity to store local reputations and, via 
ostracism and other penalties, the capacity to deter potential cheaters. 
Another important property is a network's stability, a property that may 
be studied either in terms of the attitudes of the persons involved or in 
terms of some systematic interdependence between attitudes due to 
positions in a network. In the first case, a network is stable if the attitudes 
of persons towards oilier members of a network do not produce 
psychological tension, as would a situation where the friends of my 
friends are actually my enemies. In the second case what matters is how a 



property of the positions in a network, such as the property of being 
nearby (as defined by the number of intermediary links) has effects on 
the people occupying those positions, making them more likely to adopt 
similar attitudes towards third parties. 26 Density and stability, in turn, 
may endow a community with a high degree of solidarity. This is also an 
emergent property to the extent that one and the same degree of 
solidarity may be compatible with a variety of combinations of personal 
reasons and motives; some members may be motivated by the feelings of 
togetherness which getting involved in the affairs of the community 
produces in them, others by altruism, and yet others by strict calculations 
of reciprocity. 

Components performing a material role in these assemblages include, 
in addition to the physical bodies of the people involved, the time and 
energy they must spend in maintaining relationships, two resources 
which are always in short supply and which limit the number of friends 
and contacts a person can have. The maintenance of relations involves 
more than just having frequent conversations. It also includes the 
exchange of physical aid, like taking care of other people's children, and 
of emotional support, such as giving advice in difficult situations. As it 
happens, there exists in many communities a division of labour in this 
regard with women tending to perform a disproportionate amount of the 
work involved in the maintenance of relations. 27 Components playing an 
expressive role include a large variety of nonlinguistic displays of 
solidarity and trust. Certain routine acts, such as having dinner together 
(whether on a daily basis or on special holidays) or going to church (and 
other collective rituals) serve both to express solidarity and to perform 
maintenance tasks. 28 Other acts, such as the sharing of adversity, as 
happens during a strike in a workers' community, or the demonstrated 
willingness to make sacrifices for the community, both express and build 
trust. The important point is that when it comes to express solidarity, 
actions speak louder than words. Expressive components also include any 
items capable of serving as a badge of identity. The very act of using the 
particular dialect of a language spoken in a given community, for 
example, expresses the fact that the user belongs to that community, a 
display of pride of membership which coexists with whatever linguistic 
information is communicated by words. 22 

Interpersonal networks are subject to a variety of centripetal and 
centrifugal forces that are the main sources of territorialization and 
deterritorialization. Among the former the most important is the 




existence of conflict between different communities. Conflict has the effec^H 
of exaggerating the distinction between 'us' and 'them', that is, ffl 
sharpens the boundaries between insiders and outsiders. While higW 
density itself transforms networks into enforcement mechanisms, th J 
presence of conflict increases the activities dedicated to policing 
community's borders, not only the physical boundaries of a neighbour.* 
hood or small town, but the degree to which a community controls its 
members' behaviour and promotes internal homogeneity. In othei 
words, conflict sharpens the identity of a community. This implies that 
solidarity cannot always be viewed as a desirable property since in the 
presence of conflict it results in practices of social exclusion and the 
placing of constraints on members' autonomy which greatly reduce their 
scope to be different. 50 Examples of centrifugal forces include any process 
that decreases a network's density, such as social mobility and 
secularization. Social mobility weakens links by making people less 
interdependent and by promoting a greater acceptance of difference 
through less local and more cosmopolitan attitudes. Secularization 
implies, among other things, the elimination of some of the rituals 
which, like churchgoing, are important to the maintenance of traditional 
solidarity. Transportation and communication technologies are other 
sources of deterritorialization that reduce or eliminate co-presence (i.e. 
they create dispersed interpersonal networks). Geographical dispersion 
demands that community members be more active in the maintenance of 
links, given that connections will tend to be wider and weaker and that 
readymade rituals for the expression of solidarity may not be available. 51 

There are a variety of roles that linguistic components play in these 
assemblages, an important example of which are the shared stories and 
categories that emerge as part of conflict between two or more 
communities (i.e. the narratives of 'us' versus 'them'), as well as the 
mostly stereotyped ethnic or racial categories used in them. As the f 
historical sociologist Charles Tilly has argued, these stories concentrate on 
unified space and time settings and on actors with clear motivations and j 
fixed attributes, and therefore do not really capture the actual causal 
structure of a given conflictive situation, particularly one that has lasted a 
long time. These narratives tend to leave out anything related to 
collective unintended consequences of intentional action, any process of 
accumulation or concentration of resources that is too slow to be detected 
by direct experience, as well as any social effects mediated by the physical 
environment. 32 But the role these stories play in the assemblage is not 

representing the facts but rigidifying the identities of the conflicting 
parties, the narratives being part of a process of group boundary 
instruction. In the case of ethnic communities, for instance, the 
enforcement of identity stories and categories occurs chiefly at the 
boundary. As Tilly writes, 'You can be more or less Muslim, even to the 
point when other Muslims deny your Muslimness, yet at the boundary 
with Jews you still fall unmistakably into the Muslim category.' 33 In the 
lerminology of assemblage theory, stories of conflict (and the categories 
lor insiders and outsiders associated with them) serve to code and 
consolidate the effects of territorialization on interpersonal networks. 

Speaking of conflict between communities already implies that, like all 
assemblages, interpersonal networks exist in populations. Interactions 
among members of these populations may sometimes lead to the 
formation of political alliances or coalitions among communities, 
alliances being the paradigmatic case of relations of exteriority in the 
social realm. 34 In some cases, alliances lead to the emergence of larger- 
scale entities such as social justice movements. In Tilly's view a social 
movement is composed of at least two collective actors, each comprising 
one or several allied communities with well-defined boundaries 
sharpened by conflict. One of the communities (or coalition of 
communities) must be attempting to correct a wrong or to gain a right 
of which it has been unjustly deprived; the other one is there to rival the 
claims of the first, that is, to defend advantages which would be 
threatened by its success. In other words, a movement typically breeds a 
countermovement, both of which should be considered component parts 
of the overall assemblage. In addition, the assemblage must include at 
least one governmental organization defined by its control over law- 
enforcement and military resources. The aggrieved community's goal is to 
achieve recognition as a valid interlocutor on the part of the govern¬ 
mental organization, that is, to be treated as a legitimate maker of 
collective claims, a goal that must be achieved against strong opposition 
from the countermovement. As Tilly writes: 

Claim making becomes political when governments - or more 
generally, individuals or organizations that control concentrated 
means of coercion - become parties to the claims, as claimants, 
objects of claims, or stake holders. When leaders of two ethnic factions 
compete for recognition as valid interlocutors for their ethnic category, 
for example, the government to which interlocutors would speak 




inevitably figure as stake holders. Contention occurs everywhere, but 
contentious politics involves governments, at least as third parties. 35 


Tilly discusses how the means through which political claims are made 
underwent a dramatic transformation in Great Britain between 1750 and 
1850. Claim-making moved away from machine-breaking, physical^ 
attacks on tax collectors, and other forms of direct action, towards the 
very different set of expressive displays characteristic of today's move¬ 
ments, including 'public meetings, demonstrations, marches, petitions, 
pamphlets, statements in mass media, posting or wearing of identifying 
signs, and deliberate adoption of distinctive slogans'. 36 The new 
'repertoires of contention', as he calls them, play the main expressive 
role in these assemblages. During the Industrial Revolution and after¬ 
wards, an aggrieved community (or coalition of communities) had to 
express that it was respectable, unified, numerous and committed, in short, 
that it was a legitimate collective maker of claims in the eyes of both its 
rivals and the government. 37 Of course, the possession of these properties 
may be expressed linguistically. Numerousness, for example, may be 
expressed by publishing a statement about the quantity of supporting 
members, but it will be displayed more convincingly by assembling a very 
large crowd in a particular place in town. Respectability may also be 
expressed in linguistic form, but it will be displayed more dramatically if a 
large crowd manages to stage a peaceful and ordered demonstration. 
Linguistic statements about the degree of unity in a coalition are easy to 
make, but for the same reason unity will be expressed more forcefully by 
concerted action and mutual support. 

The change in contention repertoires during the eighteenth and 
nineteenth centuries implies that some component parts switched from a 
material to an expressive role. But there are other material components, j 
Given that expressing respectability, numerousness, commitment and 
unity simultaneously is not an easy task - having numerous members 
makes presenting a unified front more problematic, for example - a large 
investment of energy on the part of organizers to hold the movement 
together is required. As Tilly writes, the 'actual work of organizers 
consists recurrently of patching together provisional coalitions, negotiat¬ 
ing which of the multiple agendas participants bring with them will find 
public voice in their collective action, suppressing risky tactics, and above 
all hiding backstage struggle from public view'. 38 In addition, given the 
fact that a government organization is always part of these assemblages, 



the list of components performing a material role must include the 
weapons, anti-riot gear and the physical control of demonstrators by 
police and army forces. The variety and concentration of means of 
coercion is an important component because the willingness and ability 
of government organizations to use them varies with circumstances, and 
this variation affords rival communities different opportunities and risks: 
a war may have just broken out and the government might need to 
recruit members of one of the communities, or, on the contrary, a war 
may have just ended and the exceptional government controls during 
wartime may have been relaxed, promoting a wave of deferred claims; 
the war may have been won or lost, increasing or decreasing the 
bargaining power of the governmental organizations and hence the 
chances of the collective actor's claims to be successfully heard, and so 
on. 39 

Questions of territorialization are directly linked to the changes in 
repertoires of contention. When direct expression of discontent was 
dominant the goals of a particular movement were more local and short¬ 
term. The switch to the new repertoire implied a move towards more 
strategic, long-term goals and this, in turn, involved the creation of 
enduring organizations to solidify gains and concentrate resources. In 
other words, accompanying the switch to the new expressive repertoire 
there was a simultaneous change in the type of collective entity that 
promoted those claims, from authorized communities to the specialized 
association, of which unions and other worker organizations are only one 
example, 40 These organizations played a key role in stabilizing the 
identity of a movement. But there are other processes that may change or 
destabilize this identity, forcing participants to invent new strategies and 
even to redefine their struggles. Among them are what Tilly refers to as 
protest cycles. These involve mutually stimulating dynamics (positive 
feedback) in which 

fone collective actors's] successful claim making tends to stimulate 
new demands on the part of other actors. That happens because some 
actors recognize previously invisible opportunities, others emulate 
newly devised means of action, and still others find themselves 
threatened by the newcomers. Expansion of claim making occurs ... 
up to the point where rivals either establish themselves, rigidity their 
positions, exhaust their energies, destroy each other, or succumb to 
state repression called forth by those whose interests claim making 



threatens ... Over such a cycle, early phases multiply innovations itj 
collective action, create relatively open spaces for new collective 
experiments, and thus give the impression of a total break with the 
past. During later phases, more moderate claimants withdraw from the 
public arena, leaving more radical and marginal activists increasingly 
isolated and vulnerable. Each large cycle of this kind leaves its traces in 
the political system: formation of new groups, alteration of relations 
between citizens and public authorities, renewal of public discourse, 
and creation of new forms of collective action. 41 

There is, finally, the question of the effect of linguistic components onj 
these assemblages. Tilly discusses the crucial role played by general terms! 
designating social categories. Given that prior to a conflict a particular social* 
group may have already been classified by government organizations! 
under a religious, ethnic, racial or other category, one of the goals off 
social movements is to change that classification. But the reason such a| 
change is important for the members of a given movement is not because 
categories directly shape perception (as social constructivists would have 
it) but because of the unequal legal rights and obligations which are attached 
by government organizations to a given classification, as well as the 
practices of exclusion, segregation and hoarding of opportunities which 
s*rt people out into ranked groups.' 12 Thus, activists trying to change a 
given category are not negotiating over meanings, as if changing the 
semantic content of a word automatically meant a real change in the 
opportunities and risks faced by a given social group, but over access to , 
resources (income, education, health services) and relief from constraints, j 
In short, struggles over categories are more about their legal and 
economic significance than their linguistic signification. 

This concludes the assemblage analysis of social justice movements. 
But there are other large social entities that are also composed of 
coalitions of networked communities, and whose identity is also shaped 
by conflict with other such groups as well as by their relations to 
government organizations: social classes. To speak of classes is to say that 
the population of networks inhabiting a particular country have 
differential access to a variety of resources and are unevenly exposed to 
a variety of constraints. In other words, the existence of social classes 
presupposes that there are processes taking place in populations of 
networks that sort them out into ranks in such a way that the persons 
composing those networks are born with different life opportunities and 



risks. To speak of ranked distributions of networks, however, is not to 
imply that the ranking is simple as in a 'society' neatly divided into upper, 
middle and lower classes. This, as Tilly argues, misrepresents the 
complexity of the relations of inequality between groups. 43 While the 
location of a network in a given hierarchical distribution of resources does 
create a set of shared interests for the persons composing the network, 
organizations are needed in many cases to focus collective attention on 
those common interests and give them a more coherent expression, as 
well as to serve as instruments of collective action when pursuing those 
interests in order to extract new rights from the government. These 
organizations, when they exist, must also be considered part of the 

The contemporary sociologist who has done the most empirical work 
in the study of these resource distributions is Pierre Bourdieu. In 
Bourdieu's view, the asymmetrical degree of access and command over 
resources acts as a force that differentiates a population of persons sorting 
them into ranked groups. Unlike older theories of social classes, Bourdieu 
does not limit his analysis to economic resources, and hence does not 
view classes solely in terms of income distributions or in terms of control 
over the means of production. To financial and industrial resources he 
adds cultural ones, such as possessing a general education or specialized 
technical knowledge, as well as owning the diplomas, licences and 
credentials needed to profit legitimately from such knowledge. This 
distinction corresponds, roughly, to the one between material and 
expressive resources in assemblage theory. In addition, Bourdieu 
emphasizes the relations that arise between positions in these distribu¬ 
tions. Examples of such relations are being above, below, or between, that is, 
the relations that exist between ranked groups. They also include 
proximity, not spatial proximity but the relation that exists between two 
groups with similar command over economic and cultural resources 
wherever they happen to be located geographically. These and other 
relations he views as relations of exteriority. 44 

The main empirical finding that must be explained, according to 
Bourdieu, is the statistical correlation between, on the one hand, positions 
in resource distributions and, on the other, a more or less coherent life¬ 
style, a term which includes both material and expressive components, the 
goods and services a given group tends to own or purchase: the set of 
manners and bodily postures it tends to exhibit; the political and cultural 
stances it tends to take; and the activities it tends to engage in within a 





whole variety of historically differentiated fields (such as economics, 
politics, religion, art). In other words, what needs to he accounted for is the 
specific mapping between a space defined by differential control over 
resources and the various fields of activity, position-takings and styles, 
Bourdieu's explanation *f the observed statistical correlations is that 
different sets of objective opportunities and risks condition the day-to-day 
practices of groups leading to the development of a durable set of 
dispositions, tendencies to behave in certain ways and to display certain 
aspirations. Considering that both habits and skills, two of the components 
of subjectivity in assemblage theory, are dispositions, most of Bourdieu's! 
ideas would seem to be ontologically compatible with the assemblage! 
approach. But there is a major incompatibility that arises due to hisj 
particular conceptualization of that set of dispositions, a set that he refers tof 
as a habitus. Bourdieu endows this habitus with a high degree of 
automatism, to the extent that all differences between the motivations 
behind social behaviour (such as the difference between causes, reasons 
and motives) disappear. As he writes: 

If a very close correlation is regularly observed between the 
scientifically constructed objective probabilities (for example, the 
chances of access to a particular good) and agent's subjective 
aspirations ('motivations' and 'needs’), this is not because agents 
consciously adjust their aspirations to an exact evaluation of their 
chances of success, like a gambler organizing his stakes on the basis of 
perfect information about his chances of winning. In reality, the 
dispositions durably inculcated by the possibilities and impossibilities, 
freedoms and necessities, opportunities and prohibitions inscribed in | 
the objective conditions (which science apprehends through statistical ; 
regularities such as the probabilities objectively attached to a group or r 
class) generate dispositions objectively compatible with these condi¬ 
tions and in a sense pre-adapted to their demands. The more 
improbable practices are therefore excluded, as unthinkable, by a 
kind of immediate submission to order that inclines agents to make a 
virtue out of necessity, that is, to refuse what is anyway denied and to 
will the inevitable. 45 

Bourdieu does not deny that, on occasion, people do make deliberate 
choices, or that sometimes they may engage in consciously matching 
means to ends. But far from constituting exceptions to the automatism of 

the habitus, it is the latter that determines when and where such 
exceptions are allowed. The habitus then becomes a master process that 
makes possible the free production of all the thoughts, perceptions, and 
actions inherent in the particular conditions of its production - and only 
those'. 46 It is not necessary to follow Bourdieu in this regard. His 
empirical observation that members of a particular class tend to display 
the same habitus may be accommodated without introducing a master 
process. We may agree, for example, that the class into which we are born 
possesses its own habits, which are regularly transmitted to new 
generations, and that it has access to special training to develop unique 
skills, a privilege that can also be handed down and preserved in a 
straightforward way. This would account for the relative homogeneity of 
a defining set of habits and skills without assuming an 'immediate 
submission to order’. Indeed, in the assemblage approach submission or 
obedience cannot be taken for granted and must always be accounted for 
in terms of specific enforcement mechanisms. The density of the 
networks structuring certain communities can be such a mechanism, as 
can be the more analytical enforcement practices of modern organiza¬ 
tions to be discussed in the next chapter. 

The main theoretical function of the habitus, that of mapping positions 
in a space of resource distributions to a space of possible life-styles, must 
also be modified. Bourdieu conceives of this space of positions as an 
abstract social space defined by two dimensions, which he calls 'economic 
capital' and 'cultural capital'. However, resource distributions never exist 
in an abstract space but are always intimately related to concrete social 
entities such as interpersonal networks and organizations. Not only are 
many resources (such as solidarity or legitimacy) emergent properties of 
these entities, but resources that have a different origin (natural resources 
like oil or coal; technological resources like machines and processed 
materials; cultural resources like diplomas or licences) are either 
controlled by organizations or produced by them. Indeed, some of the 
ranking or sorting processes that maintain the differential access to 
economic and cultural capital are resource dependence relations that exist 
not between people but between institutional organizations. 

One may wonder why a theorist of the stature of Pierre Bourdieu can 
commit himself to the existence of such an unlikely master process like 
the habitus. In what worldview, we may ask, can such a move make 
sense? The answer is not hard to find. Bourdieu believes in the 
linguisticality of experience, and therefore, he believes that all that 





needs to be accounted for is the construction of subjective experience 
through linguistic categories. 47 This is something that the notion of the 
habitus, as a set of classificatory schemes for both perception and action, 
can do quite well. That such an important author can be led astray by the 
neo-Kantian approach emphasizes the need not only for a different 
theory of experience (such as the Humean model 1 used to start this 
chapter) but also for a different conception of the role of linguistic 
categories in social processes such as that developed by Charles Tilly. 

As I said before, stories and categories play a boundary-defining role in 
Tilly's view, but these are real group boundaries not phenomenologically 
experienced borders. Tilly urges us to focus not on the linguistic label fora | 
category but on the outcome of sorting practices in a given population, | 
that is, on the practices of inclusion and exclusion that produce I 
concretely bounded groups. In other words, struggles over categories % 
are about real boundaries separating groups with differential rights and j 
obligations, boundaries that must be enforced through a variety of f 
nonlinguistic physical interventions, from imposed segregation on certain j 
neighbourhoods to forced migrations or reallocations of entire commu- * 
nities. Enforcement of categorical boundaries may also involve a variety | 
of subtler but nevertheless effective means of selectively including or I 
forcibly excluding members of certain categories from formal positions in J 
organizations. An important example of this is the matching of I 
traditionally defined categories with those created internally by economic | 
organizations. Thus, a set of stereotyped beliefs about an ethnic group, | 
widely dispersed in a population, may be matched to job categories as : 
defined by a specific commercial or industrial organization, excluding ' 
members of that group from some positions and forcing them into 
others. 48 This matching of external and internal categories is important 
because, as Tilly argues, the durability of the inequality between groups 
may be less a matter of racist, sexist or xenophobic categories than about 
the way in which these categories affect the very design of an 
organization's formal structure of roles and positions. 49 

In conclusion, we may conceptualize social classes as assemblages of 
interpersonal networks and institutional organizations. Both the net¬ 
worked communities and the organizations in which their common 
interests crystallize must be thought as having differential access to 
resources, some playing a material some an expressive role, as well as 
possessing a distinctive life-style composed of both material and 
expressive elements. A variety of practices of exclusion and inclusion 

perform the main territorializing work, while linguistic categories code 
the result of such a territorialization, consolidating the identity of a class. 
These identifying boundaries, however, must be seen as contingent and 
precarious. Social mobility, for example, can act as a deterritorializing 
process blurring the borders between classes, and technological innova¬ 
tion. by introducing novel resources, may further differentiate each class 
into several conflicting groups. Thus, we may accept that a population of 
networked communities is sorted out into social classes without having to 
agree that these classes form a simple hierarchy except in territorial states 
in which classes are relatively small and undifferentiated. 

Finally, as in the case of social justice movements, not only the 
conflicting communities must be taken into account but also the 
government organizations to which they must address their claims and 
lobbying efforts, since it is by extracting rights from such organizations 
that a given position in a resource distribution may be improved. This 
implies not only that we have a good assemblage account of political 
organizations, entities possessing an authority structure irreducible to 
network linkages, but also an account of the larger assemblages, like a 
federal government, that political organizations may form. Thus our 
ontological analysis must continue upwards to reach these larger scales 
without introducing any illegitimate entities. This is the task to be 
undertaken in the following chapter. 


Organizations and Governments 

Historically, institutional organizations have adopted many different 
forms. Even if we narrow our temporal frame of reference and focus only 
on the last two or three centuries there is still a great variety of; 
organizational forms, ranging from relatively decentralized bazaars and; 
market-places to centralized military and governmental bureaucracies. 
For the purpose of analysing the ontological status of these social entities, 
however, it is not necessary to confront this historical variety at the 
outset. Our task will be greatly simplified if we concentrate our analysis 
on those organizations involved in the imperative coordination of social 
action. But even focusing on the subset of organizations that use 
commands (as opposed to prices) to coordinate collective activity still 
leaves a very large variety of forms. We can further simplify an 
assemblage analysis if we concentrate on what all these organizations 
share in common: an authority structure. We can then separate those 
elements that play an expressive role, that is, those components that i 
express the legitimacy of the authority, from those playing a material role, | 
those involved in the enforcement of obedience, without worrying about 5 
the components that vary from one hierarchical organization to another, 
from factory equipment and weapons to corporate logos and military 
uniforms. These may be added later when making an assemblage analysis 
of concrete organizational forms. 

Max Weber, who may be considered the founder of organization 
theory, distinguished three types of authority structures according to the 
source of their legitimacy. Imperative coordination of social activity can 
occur, according to his classification, in a continuum defined by three 



extreme forms (or three 'ideal types') and their mixtures. One extreme 
form is exemplified by a perfectly efficient bureaucracy, in which a 
complete separation of position or office from the person occupying it has 
been achieved. 1 In addition, the sphere of competence of the incumbent 
must be clearly defined by written regulations and may demand 
specialized technical training tested by official examinations. Finally, 
the positions or offices must form a hierarchical structure in which 
relations of subordination between positions (not persons) are clearly 
specified in some form of legal constitution. Weber refers to this extreme 
form as 'rational-legal' to capture both the constitutional and technical 
aspects of its order, and to indicate that obedience is owed to the 
impersonal order itself, that is, that legitimacy rests on both the legality 
and technical competence of claims to authority. 2 

A second extreme form is exemplified by religious organizations or 
monarchical governments in which positions of authority are justified 
exclusively in terms of traditional rules and ceremonies inherited from 
the past and assumed to be sacred. Even in the rare case where the role of 
past precedent is breached to allow for the introduction of a novel piece of 
legislation (or other organizational change) the latter is justified by 
reinterpreting the sacred history, not by pointing out its future functional 
consequences. Unlike the previous type, a full separation of position from 
occupant does not exist, the leader or chief enjoying a sphere of personal 
prerogative within which the content of legitimate commands is left open 
and which may become quite arbitrary. Weber refers to this extreme form 
as traditional' given that voluntary submission is not to an impersonal 
order but to a sacred tradition as personified by a leader. 3 Finally, Weber 
singles out another extreme form of imperative coordination in which 
neither abstract legality nor sacred precedent exist as sources of 
legitimacy. Routine control of collective action on either basis is 
specifically repudiated by an individual who is treated by followers as a 
leader by virtue of personal charisma. Historically, the kinds of 
individuals that have played this role have ranged from 'prophets, to 
people with a reputation for therapeutic or legal wisdom, to leaders in the 
hunt, and heroes of war'. 4 This organizational type displays the least 
degree of separation of office from incumbent, and is referred to as 

Weber's classification is useful for a variety of reasons. First of all. any 
given population of organizations, even today, will tend to display a 
heterogeneous composition of authority structures approximating the 




extreme lorms. Thus, a monarchical traditional government may coexia* 
in one and the same territorial state with modern bureaucratic agencies* 
and with a variety of sectarian groups led by charismatic leaders* 
Secondly, and more importantly, many organizations tend to be mixture^ 
of different authority forms, that is, they will lie somewhere in the middl^ 
of the continuum rather than close to its extremes. Weber himseli 
discusses such mixtures in contemporary organizational arrangements,! 
like a bureaucracy commanded by an elected official who, unlike caree* 
bureaucrats appointed on the basis of their technical knowledge, may 
have been elected on the basis of personal charisma or traditional custom. 
Moreover, bureaucratic agencies whose legitimacy derives from success¬ 
fully matching means to ends also have a tendency to transform means 
into ends: that is, they tend to display a formalistic and ceremonial 
adherence to rules and procedures viewed as ends in themselves. 5 

Despite the coexistence of the three authority structures in some* 
contemporary territorial states, on the other hand, the last 200 years have 
witnessed the propagation of the rational-legal form throughout the 
organizational populations inhabiting most modern territorial states, it 
not in its extreme form then at least in mixtures dominated by this form. 
This makes this assemblage - in which the relations of exteriority 5 
between components are exemplified by a contractual relation through 
which some persons transfer rights of control over a subset of their 
actions to other persons - particularly important. Moreover, it is only in 
this type of authority structure that organizational resources are 
associated with a position not with the person occupying it. This strict 
separation results in an assemblage with clear-cut emergent properties in 
which an explanation of the behaviour of the organization does not need 
to include a description of the personal characteristics of the leaders, or in 
which such a description would be causally redundant. With a full 
separation of office from incumbent the organization itself may be 
considered a goal-oriented corporate actor. As the sociologist James j 
Coleman puts it, 'these entities, viewed from the outside, may be* 
regarded as actors, no less than individuals are. Nevertheless, from the * 
inside, they may be characterized as authority structures.' 6 

As assemblages, hierarchical organizations possess a variety of 
components playing an expressive role. Some of these are linguistic, 
such as beliefs in the legitimacy of authority, but many are not. In the 
traditional type, for example, there are many elements of rituals, like 
their choreography in space and time, that express legitimacy simply by 

conforming to past usage. In the charismatic type, it is the behaviour of 
the leader that expresses legitimacy, in the sense that he or she is obliged 
to express a strong character in one eventful situation after another. In 
the rational-legal type it is the very fact that procedures work in a 
technical sense: that is, that they regularly produce the desired outcome 
that expresses their legitimacy. On the other hand, given that sometimes 
it is not easy to evaluate whether a procedure really works, even in the 
most technical organizations the concept of 'rationality' may be used in a 
purely ceremonial way. It will all depend on the extent to which the 
quality of the outcomes (goods or services) of a technical process can be 
easily evaluated. The more complex the outputs and production 
processes, the more uncertain the evaluation, and the less clear the 
technical expression of legitimacy. In these circumstances it makes sense 
for organizations, when documenting and justifying their efficiency to 
other organizations, to stick to ceremonial 'rituals of rationality' to buffer 
themselves from criticism. 7 In the manufacture of mass-produced goods, 
for example, the technical aspect is strong and the ceremonial relatively 
weak, but in mental health clinics, legal agencies and schools, the 
evaluation of outputs may become largely ceremonial, particularly when 
these organizations express their legitimacy to government agencies 
issuing licences or permits.® 

On the other hand, and regardless of the mixture of technical and 
ceremonial components, the daily following of commands by members of an 
organization is itself a direct expression of legitimacy. In other words, 
displays of obedience, when observed by other members, directly assert 
the legitimacy of authority, while acts of disobedience directly challenge 
it. Observed disobedience, particularly when it goes unpunished, is 
detrimental to the morale of a group of subordinates. In the rational-legal 
form, where subordinates surrender rights of control expecting collective 
benefits which then translate into personal reward, disobedience 
endangers this beneficial technical outcome. In the traditional form, 
where subordinates give up control on the basis of sacred precedent, 
disobedience challenges the validity of that precedent. Thus, punishing 
disobedience in order to make an example of the transgressor is necessary 
in all authority structures, and to this extent punishment may be said to 
play an expressive role. Punishment, however, is a component that plays 
multiple roles. If the organization in question spends time deliberating 
questions such as how to fit a type of punishment to a type of infraction, 
this meshing of categories will involve linguistic components. And then. 



of course, there is torture and physical confinement, two forms of ! 
punishment with a clear material aspect. 

Like all social assemblages the material role in organizations is first and 
foremost played by human bodies. It is these bodies who are ultimately 
the target of punishment. But punitive causal interventions on the 
human body are only the most obvious form of enforcement of authority. 
Other enforcement techniques exist, particularly in the rational-legal 
form: a set of distinctive practices involved in monitoring and disciplining 
the subordinate members of, and the human bodies processed by, 
organizations. Speaking of the rational-legal form of authority, Michel 
Foucault discusses how the legitimacy of this form evolved as lawyers and 
legal scholars elaborated justifications for the contractual relations at the 
basis of voluntary submission, but also how these legitimating discourses 
had to be complemented by a nondiscursive, disciplinary component, 
which had quite different origins, not in judicial or legislative organiza¬ 
tions but in military ones. Both components converged in the Napoleonic 
state, the foundations of which, as Foucault writes, 

were laid out not only by jurists, but also by soldiers, not only 
counselors of state, but also junior officers, not only the men of the 
courts, but also the men of the camps. The Roman reference that 
accompanied this formation certainly bears with it this double index: 
citizens and legionnaires, law and manoeuvres. While jurists or 
philosophers were seeking in the pact a primal model for the 
construction or reconstruction of the social body, the soldiers and 
with them the technicians of discipline were elaborating procedures 
for the individual and collective coercion of bodies. 9 v 


These coercive procedures involve, first of all, a specific use of ' 
physical space and time. Human bodies must be distributed in space to j 
avoid unruly concentrations and to facilitate monitoring. Every 
subordinate must be assigned a definite place, whether a fixed location 
at an office or a position in a production line, so that observation of 
compliance can be routinized. The model for this analytical use of space 
was the military camp where 'the geometry of the paths, the number 
and distribution of the tents, the orientation of their entrances, the 
disposition of files and ranks, were exactly defined'. 1 * A similarly strict 
partitioning of time was performed, in which working rates were 
established, occupations imposed, cycles and repetitions regulated. 

While the use of timetables to forbid the wasting of time may be of 
monastic origin, the definition of training procedures in well-defined 
time sequences, punctuated by tests and examinations, owes much to 
military efforts to increase the efficiency of armies through the 
imposition of obligatory rhythms or 'manoeuvres'. 11 

To this strict spatial and temporal partitioning we must add ceaseless 
inspection and permanent registration to the list of components of the 
assemblage playing a material role. 12 'Permanent registration' is the term 
used by Foucault to refer to the creation and storage of records of the 
behaviour and performance of soldiers, students, medical patients, 
workers and prisoners, as a means to enforce regulations. These 
permanent records are a relatively recent historical phenomenon, a few 
centuries old at most, so an important task for the historian is to locate the 
turning point at which the threshold of description (the minimum of 
significance which a piece of information must have to be worthy of 
archiving) was lowered so as to include common people and not just the 
sacred or secular figures of the great legitimizing narratives. As Foucault 
argues, from the eighteenth century on, the 'turning of real lives into 
writing is no longer a procedure of heroization; it functions as a procedure 
of objectification and subjection'. 13 The information suitable for these 
permanent records was, in turn, the output of a variety of new forms of 
examination: from the visual inspections of patients by doctors to assess 
iheir state of health, to the tests administered to students to measure the 
degree of their learning, to the questionnaires given to soldiers to be 
recruited or workers to be hired. While previously a physician's visual 
inspection was irregular and relatively fast, now its duration was extended 
and its frequency made more uniform. While before a school's tests were 
nothing more than contests between students, they now slowly became a 
systematic way of determining, assessing and comparing individual 
capacities. In conjunction with permanent registration, the output of 
examinations allowed 'the accumulation of documents, their seriation, 
the organization of comparative fields making it possible to classify, to 
form categories, to determine averages, to fix nonns'. 14 

What processes stabilize and maintain the identity of these assem¬ 
blages? The spatial boundaries defining the limits of an authority 
structure are directly linked to its jurisdiction. In some cases, this 
jurisdiction ends at the walls of the physical building housing an 
organization, but in other cases they will extend well beyond them and 
coincide with the geographical boundaries of an entire city, a province or 



even a nation. The stability of these jurisdictional boundaries will depend 
on their legitimacy as well on their continuous enforcement. Any process 
that calls into question the extent of legitimate authority, such as a clash 
between organizations with overlapping jurisdictions, can destabilize 
their boundaries, and if the conflict is not resolved, compromise their 
identity. Similarly, a lack of economic, military or legal resources to 
enforce jurisdictional claims may blur organizational identity. Another 
source of deterritorialization in authority structures is crises of succession. 
Weber discusses a good example of these destabilizing events when he 
deals with the processes that transform a small sect ruled by a charismatic | 
leader into one of the other two organizational forms. Such sects are | 
particularly vulnerable to succession crises after a leader's death, given I 
the relative scarcity of charismatic qualities. The solution is to routinize the | 
succession process, either by making charisma hereditary (causing the * 
organization to become traditional) or by writing technical qualifications j 
which a leader must meet (thus, becoming rational-legal). As Weber 
writes: 'Charisma is a phenomenon typical of prophetic religious < 
movements and of expansive political movements in their early stages. 
But as soon as the position of authority is well established, and above all, 
as soon as control over large masses of people exists, it gives way to the 
forces of everyday routine.' 15 Routinization, therefore, is a crucial 
territorializing process in authority structures. 

Finally, there is the question of the role of language in these 
assemblages. The records and written examinations that enter into J 
enforcement practices are a good example of a linguistic component, but f 
it must be kept in mind that the kind of writing involved here is of the j’ 
logistic type, a very material form of writing documenting relatively simple j 
facts - about visits and dosages in hospitals, attendance and cleanliness in v 
schools - not the type of writing that lends itself to endless rounds of • 
hermeneutic interpretation, ft must be contrasted with other components 
of authority structures that play a straightforward linguistic role, such as 
the sacred texts or oral histories about origins which, in the traditional 
type, must be constantly interpreted and reinterpreted by the incumbents 
of certain roles, such as priests, or the written constitutions of bureau¬ 
cracies which, in case of conflict of interest, must also be interpreted by 
specialized functionaries such as judges. 

Additionally, and regardless of the form of authority, there is the role 
played by group beliefs, an emergent property of which is convergence into 
some kind of consensus. The coherence of group beliefs may be increased 

further if some specialized members of an organization (doctors, teachers, 
lawyers) routinely engage in arguments and discussions, and produce 
analyses and classifications, that transform a relatively loose set of beliefs 
into a more systematic entity, sometimes referred to as a 'discourse'. The 
systematicity of these sets of beliefs may influence not only practices of 
legitimization but also practices of enforcement. Thus, according to 
Foucault, the analytical use of space and time, the intensification of 
inspection, and the increased permanence and scope of records, all 
contributed to the development of more or less adequate technical 
knowledge in the case of some discourses such as clinical medicine, 
pedagogy and criminal law; knowledge that, in turn, increased the 
capacities of enforcement of those who deployed it. 16 

This completes the characterization of institutional organizations as 
assemblages. But, as 1 said above, besides an authority structure 
organizations also possess an external identity as enduring, goal-directed 
entities. As such organizations exist as part of populations of other 
organizations with which they interact, and in these interactions they 
will exercise capacities that belong to them as social actors, capacities that 
cannot be reduced to those of persons or interpersonal networks. The 
question now is, when organizations exercise their own capacities within 
a population of other corporate actors do larger wholes emerge? Or to put 
this differently, are there hierarchies and networks of organizations with 
properties and capacities of their own? The best example of a hierarchy of 
organizations is the government of a large nation-state, in which 
organizations may exist at the national, provincial and local levels, 
interacting with each other and operating within a complex set of 
overlapping jurisdictions. A good example of a network of organizations is 
a set of suppliers and distributors providing inputs and handling the 
output of a large industrial organization, and linked to each other 
through their relations to that dominant organization. 

Assemblage theory should apply to these larger entities in a 
straightforward way, given that both hierarchies and networks tend to 
display similar properties at different scales. There will be, on the other 
hand, differences with their smaller counterparts because the implemen¬ 
tation of strategic plans becomes more problematic, and the collective 
unintended consequences of intentional action becomes more promi¬ 
nent, at larger scales. The first question that needs to be answered when 
considering these larger assemblages is the kind of relations of exteriority 
that form between their component parts when their interactions are 


repeated over time. As 1 argued above, an organization becomes an actor 
to the extent that its resources (physical, technological, legal, financial) 
are linked to formal positions or offices, not to their incumbents. Most 
authors recognize the key role played by these resources but they tend to 
take for granted the actual process of their acquisition, even though this 
process is hardly automatic and it is often problematic for any given 
organization. In particular, organizations must engage in specific 
transactions with one another in order to solve the acquisition problem, 
and in so doing they may develop relations of dependence as these 
exchanges become more or less regular. 

The sociologists Jeffrey Pfeffer and Gerald Salancik have developed a 
useful approach to resource dependencies and to the capacity that one 
organization may have to affect the behaviour of another when such 
dependencies are asymmetrical. To define these relations of exteriority, 
Pfeffer and Salancik begin by focusing on a given organization and a 
given resource and determine the relative importance of the resource. 
Relative importance is measured both by the magnitude of the resource 
being exchanged as well as by its criticality. As they write: 

The relative magnitude of an exchange as a determinant of the 
importance of a resource is measurable by assessing the proportion of 
total inputs or the proportion of total outputs accounted for by the 
exchange. An organization that creates only one product or service is 
more dependent on its customers than an organization that has a 
variety of outputs that are disposed of in a variety of markets. 
Similarly, organizations which require one primary input for their 
operations will be more dependent on the sources of supply for that 
input than organizations that use multiple inputs, each in relatively 
small proportions ... [The] second dimension of importance concerns 
the criticality of the input or output to the organization ... Criticality 
measures the ability of the organization to continue functioning in the 
absence of the resource or in the absence of the market for the output. 

A resource may be critical to the organization even though it 
comprises a small proportion of the total input. Few offices could 
function without electric power, even though the utility may be a 
relatively small component of the organization's expenditures. 17 

In addition to the relative importance of a resource there is the 
question of its concentration, defined by the degree of control and 



substitutivity of the resource. Control refers to the capacity of one 
organization to determine the allocation of a resource for another, a 
capacity derived from ownership rights, easier physical access to the 
resource, or government regulations. Substitutivity, on the contrary, 
refers to the extent to which a dependent organization is capable of 
replacing a given supplier of the resource by another. The less alternative 
sources there are for a given resource the more concentrated it is. 1 ® 
Resource exchanges may, of course, be symmetrical or reciprocal, in 
which case the organizations may become interdependent. But if the 
symmetry of the exchanges is broken along both the importance and 
concentration dimensions then the controlling organizations acquire the 
capacity to influence the behaviour of the dependent ones. As Pfeffer and 
Salancik write, 'A resource that is not important to the organization 
cannot create a situation of dependence, regardless of how concentrated 
the resource is. Also, regardless of how important the resource is, unless it 
is controlled by relatively few organizations, the focal organization will 
not be particularly dependent on any of them.' 19 

Resource dependencies exist in both organizational networks and 
hierarchies. While in the latter case there are, in addition, authority 
relations allowing an organization with nationwide jurisdiction to give 
orders to another operating at a more local scale, the capacity to command 
on a regular and predictable basis will typically depend not only on the 
legitimacy of authority but also on the actual control of financial 
resources. However, for the purpose of giving an assemblage analysis of 
these larger entities it will be simpler to begin with the case in which 
legitimate authority is absent so that the only relations of exteriority we 
must deal with are resource dependencies. As mentioned above, networks 
of industrial organizations provide a good example of this case, but it is 
important to distinguish here two extreme forms defining a continuum of 
possibilities. The twoextremes may be characterized by different strategies 
for coping with resource dependencies. 

The first coping strategy involves the elimination of dependencies by 
the direct absorption of organizations through vertical integration, that 
is, by the acquisition of organizations that either supply inputs to, or 
handle outputs from, the absorbing firm. This strategy yields large 
organizations that are relatively self-sufficient and that can use 
economies of scale to become dominant nodes in their networks. 20 
Their dominant position allows them to control in a variety of ways 
those suppliers and distributors that have not been integrated. 




American automakers in the 1970s, for example, were capable of 
keeping their subcontractors in a completely subordinate position, their 
facilities rigorously inspected, their quality control procedures mon¬ 
itored, and even the quality and depth of their management dictated to 
them. 2 ' In a particular industry a few of these large organizations may 
coexist as separate entities forming what is called an oligopoly. This 
separation may be strengthened by the existence of legal impediments 
to the sharing of information among oligopolistic rivals, at least in those | 
countries where cartels are illegal. Nevertheless, a group of such rival | 
organizations may become linked to one another through indirect | 
means. Very large firms tend to have the formal authority structure 1 
referred to as a 'joint stock corporation', in which control and f 
ownership are separated, the former in the hands of professional * 
managers, the latter dispersed among many stockholders represented 
by a board of directors. Indirect links among j*int stock corporations 
may form through the process of interlocking directorates: the board of 
directors of a given corporation (belonging, for example, to the 
automobile industry) may include members of organizations such as 
banks or insurance companies who may also belong to other boards. 
The overlap in board membership indirectly links these organizations 1 
and protects them against the possibility of destabilizing events such as ' 
unilaterally triggered price wars. 22 i 

The second coping strategy involves not avoiding but benefiting from - 
resource interdependencies. This strategy yields networks of relatively ' 
small firms in which no organization is clearly dominant and in which the 
lack of economies of scale is compensated for by economies of agglomeration: 
many small firms agglomerated in the same geographical region tend to 
attract talented people who can find a variety of job opportunities there, 
producing over time an accumulation of skilled labour that, in turn, tends 
to expand the number of firms in the region. Thus, even though these 
industrial firms compete against each other they also benefit from their 
agglomeration and the common human resources this makes available to 
the entire region. 23 In addition, the absence of complete domination of 
subcontractors means that the relations between firms and their suppliers 
can involve more cooperation, in some cases forming a relation of 
'consultative coordination' in which firms do not command their 
suppliers to deliver components that meet rigid specifications but consult 
with them in the very design of a component. 24 If the American 
automobile industry in the 1970s illustrates the first strategy, some 

industrial regions in Italy, such as the well-studied case of Emilia 
Romagna, are a perfect example of the second one. In the early 1980s the 
manufacturing centre of this region consisted of about 22,000 firms, of 
which only a small fraction employed more than 500 employees, with a 
large percentage of the firms engaging in the design of ceramics, textiles 
and machine and metalworking products. 25 

The two extreme forms to which different ways of coping with 
resource dependencies give rise are rarely actualized, and when they are 
they are only approximated for a certain amount of time. Nevertheless it 
is still possible to compare mixtures dominated by one or the other 
extreme f*rm. In these comparisons it is important to include not only 
the industrial firms themselves but also a variety of other organizations, 
such as universities, trade associations and unions, since it is the entire 
assemblage that displays certain recurrent characteristics. Annalee 
Saxenian's study of two American industrial regions involved in the 
manufacture of computers, Silicon Valley in northern California and 
Route 128 in Boston, contrasts the properties of these assemblages. 
Saxenian writes: 

Silicon Valley has a regional network-based industrial system that 
promotes collective learning and flexible adjustment among specialist 
producers of a complex of related technologies. The region’s dense 
social networks and open labour markets encourage experimentation 
and entrepreneurship. Companies compete intensely while at the 
same time learning from one another about changing markets and 
technologies through informal communication and collaborative 
practices: and loosely linked team structures encourage horizontal 
communication among firm divisions and with outside suppliers and 
customers. The functional boundaries within firms are porous in a 
network system, as are the boundaries between firms themselves and 
between firms and local institutions such as trade associations and 
universities ... The Route 128 region, in contrast, is dominated by a 
small number of relatively integrated corporations. Its industrial 
system is based on independent firms that internalize a wide range of 
productive activities. Practices of secrecy and corporate loyalty govern 
relations between firms and their customers, suppliers, and competi¬ 
tors, reinforcing a regional culture that encourages stability and self- 
reliance. Corporate hierarchies ensure that authority remains cen¬ 
tralized and information flows vertically. The boundaries between and 




within firms and between firms and local institutions thus remain far! 
more distinct in this independent-firm system. 26 I 

When treated as assemblages of organizations the components of both 
the extreme forms as well as their mixtures play a variety of expressive 
and material roles. The former relate, in the first place, to the expressivity 
of organizational behaviour, in the sense that this behaviour may send 
signals about an organization's intentions to other members of the 
assemblage: intentions that may not be explicitly stated in the phrasing of 
a decision or in any policy-document derived from it. Although we may 
speak of 'interpretation of intentions' in this case, this will typically be not 
a matter of semantics (that is, of signification) but of assessments of 
strategic significance or importance. In the first extreme form, for 
example, an organization with a dominant position in the flow of 
resources can make claims on those that depend on it, demanding, for 
example, favourable terms of trade. But it can choose to express those 
demands loudly or quietly during negotiations, or to display its 
dominance in subtle or obvious ways. Conversely, an organization in a 
position of dependence expresses weakness by the very fact that it 
complies with demands. Acts of compliance imply an admission of limited 
autonomy, and this expression of weakness, in turn, may invite further 
demands from dominant organizations, since the latter can use the past 
actions of the subordinate organization as an indication of the probability 
of success that new claims on resources may have. In the second extreme 
form it is expressions of solidarity and trust that are important, since 
competition must be balanced with cooperation. Here what matters is the 
avoidance of the so-called tragedy of the commons', that is, the 
destruction of common resources due to the opportunistic actions of 
one organizational actor. Any action that signals a selfish disregard for 
communal welfare may trigger a series of such actions by others leading 
to the collapse of cooperation. To prevent this outcome there must be 
ways of making expressions of lack of solidarity part of an organization's 
reputation, and ways of making bad reputations have adverse economic 
consequences. This may involve either creating special organizations or 
taking advantage of the enforcement properties of dense interpersonal 
networks in a given region. 27 

The nonlinguistic expressivity of organizational behaviour is not of 
course completely unrelated to language since the actions of organiza¬ 
tions are closely linked to processes of decision-making taking place 



within them. The distinction between the two extreme forms that 
organizational networks may take is, indeed, a distinction between modes 
o( decision-making, more or less centralized in the case of large firms, 
niore or less decentralized in the case of interacting small ones. But in 
either case, decisions will be reached on the basis of beliefs about a 
number of different questions, such as beliefs about possible responses of 
other members of an oligopoly to a strategic move, beliefs about the 
degree to which dependent firms will comply with demands, or beliefs 
about the degree of solidarity in a network. All these beliefs are attitudes 
towards propositions and therefore involve a linguistic component. On 
the other hand, when we move beyond strategic decision-making to 
questions of the implementation of strategies, particularly when such 
implementation involves causal interventions in reality, these beliefs 
must now be related to the material components of the assemblage, that 
is, judged by the more or less adequate causal understanding of the 
relations between material resources that these beliefs embody, such as 
the causal adequacy of a particular technology relative to the properties of 
the raw materials serving as its inputs. Many of the resources that 
generate dependencies play a material role in these assemblages, from 
energy sources and industrial machinery, to everything related to 
logistics, from storage to transportation. Labour, skilled or unskilled, is 
another important material component. Money too may be considered to 
play a material role to the extent that its circulation causes other 
resources to flow. As systems ecologist Howard Odum puts it: 'The flow of 
energy makes possible the circulation of money and the manipulation of 
money can control the flow of energy.' 28 

The two extreme forms exhibit different types of territorialization and 
deterritorialization. Networks of small firms are married to the geogra¬ 
phical region where the organizations and the skilled workforce 
agglomerate. A single firm can make the decision to move elsewhere 
but only by giving up access to the reservoir of talent that has formed in 
that region over many years. In this sense, networks of interdependent 
firms can be said to be highly territorialized. Large, autonomous firms, on 
the other hand, having internalized a large number of economic 
functions, have for that reason acquired a certain freedom from 
geographical location. This mobility makes these firms highly deterritor- 
ialized even when they exist as national corporations, a deterritorializa¬ 
tion that is greatly intensified when globalization liberates them from the 
constraints of a national territory. But the fact that the boundaries of large 



self-sufficient firms are sharper than those of small interdependent ones 
points to a different form of territoriaiization, as does the fact that irt 
economies of scale the use of human resources tends to be very 
routinized and decision-making highly centralized, in economies of 
agglomeration, on the other hand, not only is skilled labour a crucial 
component, implying that the separation of planning from doing is not 
nearly as sharp, but it also tends to be much more mobile: the turnover 
rate, or the average time that a technical expert spends in a given job, 
tends to be two or three years in the case of networks of interdependent 
firms as contrasted with an entire lifetime in the case of many experts 
working for large corporations. 29 In this other sense, networks dominated 
by large independent firms are more territorialized than those linking | 
small interdependent ones. | 

While the assemblages of organizations populating Silicon Valley and 1 
Route 128 realize points near different extremes of the continuum of * 
forms they also interact with one another since many of these I 
organizations belong to the same industry. This implies that there are, | 
in addition, processes of territoriaiization and deterritorialization com¬ 
mon to both extreme forms, those involved in the stabilization of the 
identity of an entire industry. The integrating and regulating activities of 
organizations such as trade and industry associations are a key component of 
these processes. Industry associations are instrumental in leading their 
members towards consensus on many normative questions which affect I 
them collectively, particularly the setting of industry-wide technological 1 
standards. Trade associations can serve as clearing-houses for information ! 
about an industry's sales, prices and costs, allowing their members to 
coordinate some of their activities. They also reduce interorganizational 
variation by sponsoring research (the results of which are shared among 
members) and promoting product-definition and product-quality guide¬ 
lines. 30 The degree of organizational uniformity is also increased by the 
creation of behavioural norms by professional and worker associations: 
norms that may be informal and nonenforceable but which nevertheless 
help to standardize occupational behaviour, expectations and wages. 31 

An important deterritorializing factor affecting both forms is a 
turbulent environment, such as that created by a high rate of innovation 
in products or processes. What matters here is the relation between the 
rate of change inside organizations, a rate affected by a variety of sources 
of organizational inertia, and the rates of change of technologies outside 
of them, in the same industry in other countries, or in different industries 

in the same country. When considering entire industries we are less 
concerned with the ability of their member organizations to adapt (given 
enough time all organizations can adapt) than their ability to lime internal 
changes to external shocks, particularly when the external shocks become 
continuous. 32 To the extent that the capacity to track continuous shocks 
demands a collective response from an entire organizational network, the 
location of the network in the continuum of forms may determine the 
conditions of success or failure. The sharp separation of planning from 
doing characteristic of economies of scale limits the number of people in 
an organization that are involved in adapting to change, while the flatter 
hierarchies of smaller organizations and their use of skilled labour allows 
entire firms to learn from experience. In addition, the consultative 
coordination between firms and suppliers characteristic of economies of 
agglomeration may extend the benefits of learning by doing to the entire 
network. The faster the rate of innovation, the more a given network will 
benefit from the collective learning process of the small-firm extreme, 
and the more inadequate the self-sufficient approach of an oligopoly of 
large firms will become. 

I have already mentioned one linguistic component of these 
assemblages, but an equally important one is the written contracts (and 
other agreements) which organizations use as a means to mitigate the 
effects of interdependencies. As with the making of decisions, the content 
of contracts will vary depending on the predictability of the consequences 
of organizational action: the more eventful the situation in which a 
contract is produced, the more labour will be involved in the anticipation 
of consequences. In fact, contracts differ in the extent to which their 
wording needs to specify all contingencies and eventualities in advance. 
In neo-institutional economics, for example, a distinction is made 
between employment contracts and sales contracts, with the latter 
presenting more problems of contingency anticipation than the former. 
Indeed, when these problems are too great (due to dependencies created 
by specialized machinery, for example) this branch of economics predicts 
that an organization will switch from sales to employment contracts by, 
for example, purchasing a firm with which it previously dealt with in the 
market. 33 In addition to the difficulty presented by necessarily incomplete 
contracts (given limited rationality and honesty) a decision to use one or 
another type of contractual form may depend on the choice of the locus 
of contractual interpretation and enforcement. Whereas an employment 
contract can be enforced internally, and conflicts over its interpretation 


handled via arbitration, sales contracts must be enforced by courts and! 
interpretation conflicts handled via litigation. 34 1 

The fact that judicial interpretation of contractual obligations may be 
sometimes needed implies that the population of organizations compris¬ 
ing a given industry must include, in addition to firms, trade org an j za . 
tions and unions, an entire set of governmental organizations since the 
very legal definition and enforcement of property rights by governments 
creates the environment in which industrial and commercial firms 
depend to carry out their transactions. 35 Unlike industrial networks 
dominated by a few large firms, governmental organizations form a true 
hierarchy with a well-defined authority structure. In some cases, 
industrial networks may give rise to formal authority relations with the 
emergence of cartels, but these typically fail to have the capacities of a rea j 
hierarchy. In the 1870s, for example, before cartels were outlawed f n the* 
USA, some railroad companies attempted to give their network linkages a 
more hierarchical form, using their annual conventions as a legislative 
body (issuing rules and procedures) and a central office as an executive! 
organ implementing resolutions, but failed to create a judicial body 
capable of legally punishing violations of the cartel's rules. 3 * In the end, 
what mattered in those cartels were questions of solidarity among equally 
dominant firms not of the legitimacy of their authority. When it comes to 
governmental hierarchies, on the other hand, legitimacy is not only the 
main expressive component of the assemblage, it is a resource which 
governments can use to create dependencies, by granting or denying 
licences or certificates to organizations or professions. 

Before discussing how hierarchies of organizations can be handled in 
an assemblage approach I must make several disclaimers. First of all, it is 
impossible to discuss in the available space the large variety of forms that 
central governments have taken historically. I will therefore limit my 
discussion to those in which processes of differentiation have yielded the 
most complex forms, that is, those in which there is a clear division of j 
labour among executive, legislative and judicial organizations, an( j j n ] 
which these differentiated functions are performed simultaneously at j 
different geographical scales: the national scale, the scale of provinces or: 
states and the local scale of city government. If these complex casescan be ! 
successfully tackled, then simpler forms should present no problem. | 
Second, of all the different forms that complex central governments take I> 
will focus on the federalist form, since it displays this geographical: 
hierarchical structure most clearly. Finally, to simplify the presentation, I j 


will pick most of my examples from a single case of federalist 
government: the USA. 

In addition to warning the reader about these simplifications, I must 
make four preliminary remarks. Avoiding the use of concepts like 'the 
state' is important not only because such reified generalities are not 
legitimate ontological entities but also because such notions are too 
monolithic, that is, they fail to capture the relations of exteriority that 
exist among the heterogeneous organizations forming a government 
hierarchy. Without an adequate notion of this heterogeneity, for 
example, we could make the mistake of thinking that there is no gap 
between the formulation of public policy and its actual implementation, or that 
a government's capacities to intervene in reality are related in a simple 
way to the decisions made by some elected representatives to perform 
such interventions. But studies of the process of implementation have 
shown just how difficult it is to go from a document summarizing a goal 
to be achieved, to the process of choosing the right agencies to carry out 
the policy, to maintain the commitment and flow of funds required at 
different stages and, in general, to ensure compliance in a long chain of 
national, state and local governmental organizations with overlapping 
jurisdictions. In many cases, central policy decisions end up either not 
implemented or changed beyond recognition. Joint action by many 
governmental organizations is thus objectively complex and problematic, 
not something that can be taken for granted. 37 Of course, the complex 
relation between policy formulation and implementation may be 
interpreted as implying that the two activities form a seamless web: an 
interpretation that would bring us back to a monolithic concept. But it 
can also be modelled as a nonlinear process involving feedback, a process 
of formulation-implementation-reformulation that does not jeopardize 
the ability to assess the extent of goal attainment and the distribution of 
authority between elected and appointed officials'. 38 

The second preliminary remark expands on this last point. The 
relations between government organizations staffed by elected officials 
(that is, democratic or representative organizations) and those run by 
non-elected, career bureaucrats are problematic in a deeper sense. In 
order for bureaucracies to be run efficiently, a sharp separation between 
politics and administration is necessary: that is, the expertise of a 
professional body of bureaucrats must be isolated from the contingencies 
of the electoral process. But the more this separation is achieved, the 
greater the sense that bureaucracies are not responsive to public concerns 



as expressed in electoral outcomes. In other words, the same factors that 
promote efficiency tend to undermine legitimacy, at least in democratic 
regimes. One element of this conflict is common to many social relations 
that involve delegation of authority. In one model (the 'principal-agent' 
model) the problem is framed like this: how can employers (the 
principals) make sure that no cheating and shirking will occur if they 
have less expertise than the agents they hire and to whom they delegate 
authority? In this model the basic conflict emerges from expertise 
asymmetries, and may be applied at larger scales because neither presidents 
nor legislators (nor their respective staffs) have the specialized knowledge 
needed to assess the performance of bureaucracies. 39 But this model 
leaves out other problems that do not have counterparts at smaller scales. § 
In particular, the very expertise asymmetries that favour bureaucracies I 
may be turned against them, since in many cases (atomic power, j 
pharmaceutical products, financial processes) the industries that govern- I 
ment agencies are supposed to regulate supply them with the very jj 
technical information they need to enforce regulations. In other words, f 
regulatory agencies may become captive of special interests, that is, j 
dependent on their technical resources, further eroding their already f 
questionable legitimacy. 40 I 

The third and fourth preliminary remarks concern distinctions that are | 
crucial within assemblage theory but that are not necessarily drawn in I 
other approaches. First of all, we must distinguish between the hierarchy j 
of organizations forming a federal (or other form of) government from | 
the territorial entity such as hierarchy controls. The territorial entity j 
includes, beside government organizations, an entire population of other j 
organizations; populations of persons and interpersonal networks; cities, : 
regions and provinces; and geopolitical relations of exteriority with other j 
territorial entities. When a political revolution changes one government | 
regime by another, an autocratic regime by a democratic one, for 
example, it typically leaves untouched the previous unequal relations 
between cities, regions and provinces, not to mention the geostrategic 
position of the country relative to other countries. On the other hand, 
this distinction should be made carefully since most hierarchies of 
organizations are not really separable from the territory they govern, and 
part of what defines their identity is exercising actual control over the 
borders of that territory. Unlike interpersonal networks or institutional 
organizations which, thanks to communications technologies, may exist 
without well-defined spatial boundaries (or even in virtual form on the 


Internet) complex organizational hierarchies can hardly be conceptua¬ 
lized outside the territory they control or the resources (natural and 
demographic) associated with that territory. Nevertheless, in what 
follows I will emphasize the characteristics of the assemblage of 
organizations itself leaving the analysis of the territorial aspects for the 
next chapter. 

In addition to distinguishing the hierarchical assemblage of organiza¬ 
tions from the kingdom, empire or nation-state that it controls, it is 
important to separate for the purposes of analysis the enduring assemblage 
itself from its interactions with other organizations, with coalitions of 
networks, or with populations of individual persons. Some of these 
interactions may also yield assemblages, constituting complex political 
situations: assemblages that are large-scale counterparts of conversations 
among persons. In the previous chapter I discussed Charles Tilly's ideas 
about social justice movements as assemblages of coalitions of networks 
and government organizations acting as interlocutors. Tilly sees public 
demonstrations as large-scale conversations between a movement, a 
countermovement and the police. More generally, he writes that, whether 
in the ritual executions, processions, celebrations, and militia marches of 
the early French Revolution or the public meetings, petition drives, 
lobbying, demonstrations, and association-forming of contemporary 
Western social movements, we witness the conversational combination 
of incessant improvisation, innovation, and constraint'. 41 

Like personal conversations, in which claims to a public persona are 
made by its participants, conversations between organizations (or 
between organizations and network coalitions) also involve claim¬ 
making and collective production of identities: the identities of an ethnic 
community or of an industrial sector, for example. But like personal 
conversations, these interactions are highly episodic and do not 
necessarily change the identity of the government itself, except in the 
case of political revolutions. Also, conversations are only one example of 
a social encounter, a term that encompasses a wide variety of episodic 
assemblages - a point that applies at larger scales as well. Thus, in what 
follows, I will give an assemblage analysis of the hierarchical assemblage 
of organizations first, and then add a single example of the large variety of 
episodic assemblages it forms through its interactions. 

As in all assemblages possessing a command structure, the expressive 
role is played by those components involved in the legitimization of 
authority, while the material role is played by components involved in its 



enforcement. In the USA, for instance, there are two main sources (j 
legitimacy at this scale, the constitution and the electoral process. Thi 
constitution is, of course, a linguistic component, a binding writtei 
document specifying, among other things, the relations between 
executive, legislative and judicial organizations, as well as those between 
organizations operating within national, state and local jurisdictions. Th< 
electoral process is a nonlinguistic component endowing elected official! 
with legitimacy to the extent that its outcomes express the will of th< 
population. But the mere ceremonial conduct of elections does not, in 
fact, ensure that there will be proportional representation of the different 
groups in the electorate. There are technical features oi votin 
procedures, such as how are votes aggregated or how winners at 
selected, that impinge directly on the question of how well th 
preferences of a population are expressed in electoral outcomes, an 
hence, how representative and legitimate are the results. 

There are, for example, voting systems in which voters only have on 
vote, and in which the candidate with more votes wins (plurality voting 
systems in which voters get many votes that they can allocate in differer 
forms (approval voting); and systems in which votes determine not a yes 
or-no choice but a ranking of the candidates (preference voting). The 
capacities of these voting systems to express actual distributions of! 
collective preferences are quite different, as are their vulnerabilities to 
strategic (or tactical) voting, that is, voting not for one's real preference 
but to prevent someone else from winning. 42 Although mathematicians 
disagree on which system is best - and given that voting may be 
performed for many different purposes, there may not be a best choice - 
they all agree that plurality voting is technically the worst, so its survival 
in modern nations such as the USA may be explained by its ceremonial | 

If these were the only two sources of legitimacy, then the problem of f 


bureaucracies would be insoluble and lead to continuous crises: bureau- f 
crats are not elected officials so they cannot draw legitimacy from| 
electoral outcomes, and the constitution is mostly silent about the status! 
of bureaucracies and about the legitimacy of delegating to them: 
investigative, prosecutorial and adjudicating authority: a delegation 
which would seem to violate the doctrine of separation of powers. 43 
But there are other sources of legitimacy. When discussing Weber's 
theory of authority I mentioned that in the rational-legal form the 
technical efficiency of procedures itself is an expression of legitimacy. In 



France and England, where bureaucracies emerged prior to democratic 
regimes and were staffed with members of an elite public service, 
technical efficiency often played this legitimizing role. But in the USA the 
opposite historical sequence occurred, so that it was only in the context of 
the Great Depression of the 1930s that disinterested expertise was used as 
a pragmatic justification for the existence of bureaucracies 44 Even then, 
however, distrust of specialist knowledge (as opposed to the more 
generalist knowledge possessed by elected officials) made this a 
precarious expression. So another expression of legitimacy soon 
appeared: the fairness of the procedures used in bureaucracies, as well as 
the degree to which these procedures were standardized across all 
commissions and agencies. These questions were codified in 1946 in the 
Administrative Procedure Act. As with the fairness of voting procedures 
there are technical issues involved, so the problem is not one of 
negotiating the meaning of the word 'fair'. In the hearings conducted by 
regulatory agencies, for example, the roles of judge and prosecutor 
cannot be played by the same staff member without introducing bias. The 
Act had to, therefore, create a special group of hearing examiners isolated 
from such conflicts of interest, in order to increase the legitimacy of 
administrative justice. 45 

Much as physical punishment and confinement can be used to enforce 
authority on individual persons, military and police organizations can be 
used by central governments to secure compliance from bureaucracies and 
local officials. Systematic reliance on physical force, however, signals an 
unstable form of authority, so other material components must be added 
to these to align enforcement and legitimacy. Presidents and legislators 
have the capacity to control bureaucrats in a variety of ways: presidents 
have the power of appointment and removal of key personnel, as well as 
control of financial resources; legislators can exercise control by designing 
bureaucracies, that is, they can build incentives against cheating and 
shirking into the very legal mandates that establish the goals and legal 
form of a new agency. Careful quantitative studies based on the principal- 
agent model have shown that executive and legislative organizations not 
only have these capacities but that they actually exercise them. 46 Congress 
also has oversight committees that monitor bureaucratic efficiency, and 
the courts can perform judicial reviews to make sure that due process is 
respected in the conduct of administrative justice. 

When considering processes of territorialization it is important to 
distinguish between the identity of individual policies and the identity of 



the assemblage of organizations itself. The relative political autonomy qj 
bureaucracies is clearly not a stabilizing factor in the former case, but it is 
so in the latter. Before a merit system and a career civil service wer^ 
instituted in the USA in 1883, for example, bureaucratic offices wer* 
considered spoils to be given to the winners of an election. In that 'spoil 
era', the identity of the entire assemblage could be affected by episodii 
shifts in public opinion. But once a certain degree of insulation frorr 
politics was achieved, bureaucracies became sources of continuity an< 
long-term coherence. In a sense, since the legal mandate that brings j 
bureaucratic agency into existence may reflect policies different from 
those of currently elected officials, political insulation may provide a 
mechanism for policy integration across different administrations. 4 ! 
Given that the relative autonomy of bureaucracies is partly based on 
expertise asymmetries, a main territorializing process is the professional 
lization of civil service personnel, a professionalization that has taken| 
different forms in different countries. In France, for example, it wa£ 
closely linked to the training of civil servants in elite universities and 
polytechnics, a common educational basis that instilled an esprit de corpM 
on potential candidates. In England, it was through on-the-job trainin'* 
that expertise was passed to new recruits: a learning process that fostere 
loyalty to the office itself as opposed to its current incumbents, 48 

Among the deterritorializing processes that affect the identity of thes 
assemblages from within (as opposed to from without, as in political! 
revolutions) two stand out: coups d'etat and constitutional crises. The 
former involves a change of regime forced on central organizations byj 
other government organizations, typically military ones, or by organiza-l 
tions that have wrestled control of the army from the executive branch. At 
coup d'etat is not only destabilizing as an event. Even when it is over the! 
new incumbents will typically possess very little legitimacy (in the eyes oft 
other government organizations as well as the rest of the population) and| 
will have to resort to physical coercion as the main instrument of’ 
authority enforcement. 49 Constitutional crises can have a wide variety of.: 
causes, such as a succession crisis due to ambiguous electoral results. But) 
a crisis may involve a more complex situation in which different 
government organizations are pitted against each other. Executive ‘ 
organizations, for example, may refuse to recognize the legitimacy of 
legislative ones, calling for their dissolution, while at the same time a 
legislative body may question the legitimacy of a president's actions and 
call for his impeachment. (Something like this happened in Russia in 


j 993 .) On the other hand, the conflict may involve not two branches of 
government but organizations operating at different geographical scales, 
as when local or state governments refuse to obey central commands. In 
the nineteenth century, for example, the conflict over slavery in the USA 
proved insoluble via existing mechanisms (such as Supreme Court 
decisions), provoked the secession of eleven southern states, and had to 
he resolved by civil war, and by the constitutional amendment that 
eventually outlawed the practice. 

There is, finally, the question of the role that linguistic components 
play in these assemblages. I mentioned above the variety of means that 
executive, legislative and judicial organizations have to control bureau¬ 
cracies. Those means, however, are mostly of strategic value, being useful 
in securing overall compliance but powerless to determine specific 
outcomes, given that administrative agencies may use their relative 
insulation from politics to shape the implementation of centrally decided 
policies. Tactical means, such as the unambiguous wording of the original 
policy document (or statute), must also be used to maintain the integrity 
of policy decisions. 5 * I also mentioned the most crucial binding document 
in many countries: the basis for more or less codified forms of constitutional 
or basic law. These laws not only consolidate the identity of the 
assemblage (i.e. they perform a coding operation to complement the 
effects of territorialization), they also limit the kinds of other laws 
legislative organizations may create. These other laws vary in their degree 
of codification and in the extent to which custom and precedent may 
affect their interpretation, as in the difference between the common law 
prevalent in England and its ex-colonies, and the more systematic, less 
precedent-bound, civil law prevalent in the countries of Continental 
Europe and their ex-colonies. These and other written laws form the 
institutional environment for the economic organizations that I discussed 
before, as well as for all the other social assemblages we have considered 
so far. 51 

This brings me to the question of the more or less episodic interactions 
between hierarchical assemblages of organizations and other social 
entities. Of all the different interactions I will pick a single one, 
interactions with a population of persons, and of all the different political 
situations in which these interactions may take place I will select the 
situation created by the existence of armed conflict, whether external or 
internal. On the material side, this situation calls for both recruitment of 
People - sometimes voluntary, sometimes coerced - as well as the 




necessary taxation to pay for war. The central policies in which th 
goals are stated (a draft resolution, a change in fiscal policy) must ta! 
into consideration resistance from the target groups, so they involve 
concessions and political dialogue. The situation may be framed ij 
resource-dependence terms: taxes and military recruits are such at; 
important resource for a government that it comes to depend on its 
population to obtain them, thereby becoming subject to its demands. Iij 
fact, according to Charles Tilly, this is exactly how modern rights o; 
citizenship came into being in Europe in the seventeenth and eighteenth 
centuries, as governments engaged in the expansion of their armies (and 
of the taxes needed to pay for them) had to bargain with the target- 
groups and yield to their demands for political participation. 52 

On the expressive side, these situations call for a variety of means 
some symbolic, some directly expressive - to strengthen the unity of a 
government and the population. The classical example is the effect that: 
the French Revolution had on the composition of armies, that is, the! 
change from mercenary to loyal citizen armies. The means used to effectp 
this change in different countries, however, varied with the existing! 
sources of legitimacy. Two of the forms of legitimacy discussed by Weber,,! 
traditional and rational-legal, have counterparts at larger scales. In some) 
countries the bonds uniting a population are inherited or come from a 
long tradition, so that the 'nation' precedes the 'state'. In others, these! 
bonds emerge from the sharing of the same laws, that is, the 'state'! 
precedes the 'nation'. 53 Countries that followed the state-to-nation path 
(such as France or England) tended to favour newly invented expressions 
of patriotism: flags, oaths, anthems, national holidays, military parades, 
official celebrations. Those that followed the nation-to-state path 
(Germany) tended towards more populist expressions, using more or 
less coherent syntheses of popular elements created by intellectual elites.: 
However, just as Weber's ideal types rarely exist in pure form, blood and 
law as sources of national unity were never mutually exclusive. Most 
countries used a mix of these two sources of legitimacy when rallying 
their populations for war. And ultimately, regardless of what combination 
of expressive means a given government used, the ultimate display of 
patriotism has always been the willingness of citizens to die for their 
country, as expressed behaviourally on the battlefield. 

The reality or threat of armed conflict is itself a powerful territorializ¬ 
ing force, making people rally behind their governments and close ranks 
with each other. Much as the solidarity binding a community may be 

[ra nsformed into social exclusion when conflict with other communities 
sharpens their sense of 'us' versus 'them', external war can transform a 
simple emotional attachment to a country's traditions and institutions 
j n to a sense of superiority relative to enemy countries and their allies. 
Loyalty, which need not involve comparisons with others, is transformed 
into hostility and xenophobia. Internal war, on the other hand, can act as 
a eleterritorializing force, either by destabilizing a government through 
constant riots and turmoil or by drastically changing its very identity, 
from one regime to another, as in successful political revolutions. Unlike 
coups d'etat, revolutions go beyond interactions between government 
organizations. The minimum assemblage, a recurrent one in past 
revolutions, includes: a population that has undergone a period of 
relative prosperity and rising expectations, followed by a period of 
deprivation when those expectations are frustrated; a struggle between 
dominant coalitions and those who challenge them; and displays of 
vulnerability by government organizations, such as a decrease in their 
enforcement capacities due to a fiscal crisis, a bad economy or a military 
defeat abroad. 54 

While for the citizens of a given country external warfare may not 
have a definite spatial dimension, in the sense that they may form 
xenophobic beliefs without a dear sense of the territorial situation of 'us' 
versus 'them', for government organizations this is not typically the case, 
unless the threat comes from terrorist organizations lacking any territorial 
base. For most of their modern history, however, governmental 
hierarchies have operated within concrete geopolitical entities, such as 
nation-states, kingdoms or empires. Moreover, international law, as it 
developed in the West after the peace treaty that ended the Thirty Years 
War in the seventeenth century, was intimately related to spatial 
questions, such as legal definitions of sovereignty within bounded spatial 
territories, and geostrategic questions defining the military opportunities 
and risks that different organizational hierarchies had to face. Thus, we 
have reached the limits of what can be analysed without reference to the 
spatial aspects of assemblages. In the following chapter I will return to the 
analysis of government organizations and of the processes that produced 
modern nation-states, once I have dealt with the spatial aspects of 
assemblages at smaller scales, from buildings and neighbourhoods to 
cities and the hierarchies and networks that urban centres form. 



Cities and Nations 

Interpersonal networks and institutional organizations may be studied* 
without reference to their location in space because communication| 
technologies allow their defining linkages and formal positions to be* 
created and maintained at a distance, but as we move to larger scales| 
spatial relations become crucial. Social entities like cities, for example, 
composed of entire populations of persons, networks and organizations, 
can hardly be conceptualized without a physical infrastructure of 
buildings, streets and various conduits for the circulation of matter and 
energy, defined in part by their spatial relations to one another. In fact, 
sociologists discovered the social relations generated by territoriality in 
the 1920s when the famous Chicago school began its studies of urban | 
contexts, viewed both as spatial localities as well as sites structured in I 
time by habitual or customary practices. 1 More recently sociologists such | 
as Anthony Giddens, influenced in part by the work of urban 
geographers, have returned to this theme, reconceptualizing social 
territories through the notion of a ‘regionalized locale'. As Giddens 

Locales refer to the use of space to provide the settings of interaction, 
the settings of interaction in turn being essential to specifying 
contextuality ... Locales may range from a room in a house, a street 
corner, the shop floor of a factory, towns and cities, to t he territorially 
demarcated areas occupied by nation-states. But locales are typically 
internally regionalized, and the regions within them are of critical 
importance in constituting contexts of interaction ... One of the 

reasons for using the term 'locale' rather than place is that the 
properties of settings are employed in a chronic way by agents in the 
constitution of encounters across space and time. fLocales can be] 
‘stopping places' in which the physical mobility of agents’ trajectories 
is arrested or curtailed for the duration of encounters or social 
occasions ... 'Regionalization' should be understood not merely as 
localization in space but as referring to the zoning of time-space in 
relation to routinized social practices. Thus a private house is a locale 
which is a 'station' for a large cluster of interactions in the course of a 
typical day. Houses in contemporary societies are regionalized into 
floors, halls and rooms. But the various zones of the house are zoned 
differently in time as well as space. The rooms downstairs are 
characteristically used mostly in daylight hours, while bedrooms are 
where individuals 'retire to' at night. 2 

Giddens' description of regionalized locales, as physical territories 
structured in time by social rhythms, lends itself nicely to an assemblage 
approach, providing his definition is augmented with the expressive 
elements with which locales and regions distinguish themselves from 
each other. The stress on rhythmic or periodic routines, however, would 
seem to present a problem. I have argued in previous chapters that, 
except in the most uneventful situations, routine behaviour must be 
complemented with deliberate decision-making in the explanation of 
social action. But when studying the effect of human behaviour on the 
form of urban components the emphasis on routine activity is justified 
because, as the historian Fernand Braudel reminds us, urban forms tend 
to change extremely slowly. A house, as he says, 'wherever it may be, is 
an enduring thing, and it bears witness to the slow pace of civilizations, of 
cultures bent on preserving, maintaining, repeating'. 3 Given this 
slowness it seems correct to emphasize those human activities that are 
so regular they have a chance to impinge on urban form in the long run, 
such as the journeys to work or journeys to shop that give cities their 
daily rhythms. On the other hand, in those cases where we witness 
historical accelerations of this slow pace we will have to add choice to 
routines since acceleration in the change of urban form typically implies 
breaks with tradition and hence, deliberate design. 

Let us now give an assemblage analysis of these regionalized locales, 
starting with individual buildings. The material role in buildings is played, 
first of all, by those components that allow them to be successful load- 



bearing structures. For buildings that are a few storeys high, the wa: 
themselves perform this task, in conjunction with columns a: 
independent beams, but large governmental, religious and corpora 
buildings must make use of more sophisticated techniques as thi 
become taller. As skyscraper designers know well, radical changes in fo: 
may be needed once a critical height has been reached, such as the use 
an interconnected iron or steel frame which, beginning in the 185 
liberated walls from their load-bearing duties transforming them in 
mere curtains. Other components playing a material role are thoi 
determining the connectivity of the regions of a building. If locales an 
stations where the daily paths of individual persons converge, the regio: 
that subdivide them must be connected to each other to allow for thi 
circulation of human bodies and a variety of other material entities.' 1 In 
simple dwelling, this connectivity is effected via doors, hallways am 
staircases shaping the flow of people, and by windows for the circulation* 
of air and light. In taller buildings, on the other hand, internal! 
transportation technology may be needed. Thus, the same decade thafj 
saw the introduction of the internal metallic frame also witnessed the] 
transformation of old mechanical lifting devices into the earlie; 
elevators, and a corresponding transformation in the vertical connectivity 
of buildings. 

Changes in connectivity, in turn, impinge in a variety of ways on the! 
social activities performed in a given locale. Fernand Braudel, for 
example, argues that the connectivity of some residential buildings in 
the eighteenth century changed dramatically at the same time that the! 
function of the rooms became more specialized, with the bedroom in| 
particular becoming a fully detached region. As he writes, the new? 
connectivity contrasted sharply with that which characterized previous! 
buildings: I 

In a Parisian town house of the seventeenth century, on the first floor, 
which was the noble storey, reserved for the owners of the house, all 
the rooms-antechambers, salons, galleries and bedrooms - opened off 
each other and were sometimes hard to tell apart. Everyone, including 
servants on domestic errands, had to go through all of them to reach 
the stairs. 5 

A hundred years later, some rooms had become public while others were 
strictly private, partly as a result of the fact that the routine circulation 

through a house was now constrained by a different distribution of doors 
and hallways. Privacy, in a sense, was created by the new regionalization 
•f these locales. In nonresidential buildings the changes in connectivity 
brought about by elevators altered the form of the circulation of 
employees, from a horizontal to a vertical form, whenever firms were 
n*t able to secure nearby buildings to accommodate a larger number of 
workers. As the urban geographer James Vance writes: 

For the financial district [the) mechanical lift was of critical 
importance, because much of the movement tended to be internal 
to a rather clearly defined group of employees in a single organization 
or in a modest number of commonly related organizations. In that 
situation the walking zone limits could be reached within a few 
adjacent buildings, as in the structures built to house a legal 
community, a medical one, or even a very large single insurance 
company ... It seems to me not at all a matter of chance that the 
earliest skyscrapers to be built, those in New York and Chicago, were 
constructed predominantly for insurance companies and were among 
the earliest buildings to be equipped with elevators. Large metropo¬ 
litan newspapers were other early entrants into the construction of 
skyscrapers, again finding a great advantage in piling large numbers of 
workers on top *f each other and thus, by elevator, being able to 
secure rapid personal communication. 6 

The introduction of internal transportation also had expressive effects. 
Thus, the apartment buildings that were constructed prior to the elevator, 
in Paris for example, displayed a clear vertical stratification in which the 
social status of the inhabitants decreased with height. After the elevator 
was introduced, this stratification of regions was reversed, with 
apartments higher up expressing increased status. 7 Other expressive 
components vary with the activities housed by the building. In the case of 
residential buildings, the distinctive furniture of their internal regions and 
the decorative treatment of walls, floors and ceilings, have often played a 
role in the marking of social-class territories. Ostentatious displays in the 
aristocratic homes of Renaissance Italy, as Braudel reminds us, were in 
fact a way of using luxury as a means of domination. But as he goes on to 
argue, this luxury was purely expressive, since until many centuries later 
it was not associated with any kind of material comfort. 8 In the case of 
public buildings, particularly important examples are cathedrals, 



churches, mosques and synagogues: locales used for worshipping] 
services, processions and religious ceremonies. These buildings musts 
demarcate a sacred territory from a profane one through the expressive! 
use of geometry and proportion. In medieval Europe, for example, the! 
overall cruciform shape, arcaded cloisters and rhythmic patterns ini 
stained-glass windows were all sacred territorial markers. No doubt, these- 
spatial expressions often coexisted with religious representations. The 
fan-vaults of some English Gothic churches, for instance, with their series 
of ribs radiating upwards, express an expansive, ascending motion well 
suited to mark a sacred territory. This physical expression, of course, must 
work in conjunction with linguistic ones (the belief that, for example, 
heaven is above the earth), but it is not reducible to them. 

What are the processes that stabilize or destabilize the identity of these 
assemblages? In the Chinese, Indian and Islamic civilizations, as well as 
among the European poor, the weight of tradition seems to have been 
almost overwhelmingly stabilizing when it comes to building techniques 
and materials, as well as the evolution of furniture and other elements of 
interior decoration. This evolution, when it took place, occurred at a 
glacial pace. The birth of fashion, on the other hand, had deterritorializing 
effects, although these were at first confined to the European rich. 
Fashion greatly accelerated the pace at which the interior and exterior 
decoration of buildings evolved, although it was not until the 1700s that 
the rate of change approximated the speed to which we have become 
accustomed today. 9 The impetus behind fashion was not just the desire to 
mark social-class territories through the way bodies and homes were 
dressed but also derived from the fact that, in Europe, aristocracies saw 
their distinguishing expressive markers constantly under threat by the 
increased social mobility of rich merchants and artisans. This resulted in a 
spiralling 'arms race' that drove change. As Braudel writes: 

f have always thought that fashion resulted to a large extent from the 
desire of the privileged to distinguish themselves, whatever the cost, 
from the masses that followed: to set up a barrier ... Pressure from 
followers and imitators obviously made the pace quicken. And if this 
was the case, it was because prosperity granted privileges to a certain 
number of nouveaux riches and pushed them to the fore. 10 


Another process deterritorializing the identity of buildings is drastic 
changes in the routines which give them a temporal rhythm. In the case of 


organizations possessing an authority structure, changes in either practices 
of legitimization or enforcement may affect the identity of locales. As new 
enforcement routines replaced old ones in the seventeenth and eighteenth 
centuries, for example, they generated a distinct regionalization and 
connectivity in the buildings of factories, prisons, hospitals and schools. As 
Michel Foucault writes, these buildings have 

an architecture that is no longer built simply to be seen (as with 
ostentatious palaces), or to observe the external space (cf. the 
geometry of fortresses), but to permit an internal, articulated and 
detailed control - to render visible those who are inside it ... The old 
simple schema of confinement and enclosure - thick walls, a heavy 
gate that prevents entering or leaving - began to be replaced by the 
calculation of openings, of filled and empty spaces, passages and 
transparencies. 11 

We can extend these remarks to other types of locales, such as office 
buildings. The bodies of bureaucrats, for example, must also be 
analytically distributed in space, pinned down to their offices, and 
separated from any activity not directly related to their jobs. 'The physical 
separation of offices', Giddens writes, 'insulates each from the other and 
gives a measure of autonomy to those within them, and also serves as a 
powerful marker of hierarchy.' 12 

The changes brought about by fashion, or by the disciplinary use of 
space, already point to the fact that buildings exist in collectivities of 
similar assemblages, since in both cases we are concerned with how new 
forms propagate over time through an entire population. These 
populations of buildings, in turn, form larger assemblages such as 
residential neighbourhoods, commercial, industrial or government 
districts, or even moral (or immoral) zones, such as red-light districts. 
What components play a material or expressive role in these larger 
assemblages? On the material side, we must list all the physical locales 
defining stations for the periodic intersection of the life paths of 
neighbours (the local square, churches, pubs, shops) as well as the 
streets providing the necessary connectivity among them. A whole 
underground infrastructure, starting with water and sewage pipes and 
conduits for the gas that powered early street lighting, was added in the 
nineteenth century, and the twentieth contributed with electricity cables 
and telephone wires. 



On the expressive side, it was the exterior of buildings, that is, th« 
decoration (or lack of it) of their facades, that defined the personality of a 
neighbourhood. Jn residential neighbourhoods where streets were! 
narrow and their layout formed a complex maze, the frontage of housed 
remained rather plain. Hence, expressive exteriors appear first in public] 
buildings. These were typically located on a central square in which the] 
surrounding space opened up vistas, that is, opportunities for unusual; 
visual experiences, and effect enhanced by a straight street leading to the] 
church, administrative building or monument. Aristocratic residential 
buildings joined public ones by the fifteenth century, as the European 
rich began deliberately to pick observable sites for the location of private 
houses. Only when enough space was left open around these buildings 
could expressive ostentation, and the interclass competition that fuelled 
it, begin to touch the external surfaces. 13 Besides opening up vistas, the 
central square of a town played another expressive role: as a centre 
determining the location of residential neighbourhoods, with proximity 
to it expressing greater social prestige. This concentric arrangement was 
characteristic of many European medieval towns, but was more prevalent 
south of the Alps. In the north, where merchants or craftsmen dominated 
their settlements, a market-place occupied the centre of the city, and 
accessibili ty to it determined the desirability of a location. This functional 
rather than social separation led to a more egalitarian form of 
expressivity, particularly in those planned towns named 'bastides' which 
were used in the late Middle Ages as a means to colonize economically 
backward areas within Europe. 14 

Next we must list the processes that sharpen the boundaries and 
increase the internal homogeneity of a given neighbourhood. The 
processes of congregation and segregation are among those that perform 
this territorializing function. As James Vance writes: 

The activities that grow up in cities show a strong tendency to come 
together in limited areas of specialization drawn into a congregation 
by the internalizing linkages among them. Whether it be the use of 
shared sources of materials, the selling to a common body of 
customers, the practice of a given religion, or the speaking of a 
particular language, the institutional practice shapes the process of 
congregation, which is internally induced and highly responsive to 
matters of scale. A few persons doing a particular thing normally 
congregate, but not in an obvious congregation. When numbers are 



increased to the point that they present an areally extensive pattern, 
then a geographical congregation is to be seen ... In contrast to a 
congregation is a similarly extensive grouping of ostensibly similar 
individuals induced by external forces. Instead of being drawn 
together, they are forced together by segregation. 15 

Commercial and industrial neighbourhoods have often been subject to 
the processes of congregation and segregation: similar crafts and trades 
have traditionally tended to congregate, while certain noxious activities 
like slaughtering have often been the target of institutional segregation. 
But residential neighbourhoods too acquire relatively well-defined 
borders, and a uniform internal composition, through these processes. 
The case of institutionalized segregation is perhaps the clearest example, 
since in this case both the boundaries and composition of a neighbour¬ 
hood are codified by law and enforced by government organizations. But 
congregation may also result in a relatively homogenous composition (by 
race, ethnic group, class, language) even when one assumes a desire by 
residents to live in a relatively integrated neighbourhood. If people who 
do not actively discriminate also prefer not to be in the minority, whether 
relative to their immediate neighbours or relative to their overall 
proportion in the neighbourhood, there will be critical thresholds in 
the composition of a neighbourhood beyond which a chain reaction takes 
place causing a flight away from the locale by one of the groups. 

Important examples of processes of deterritorialization are increased 
geographical mobility and the effect of land rents on the allocation of uses for 
a particular neighbourhood or district. As the sociologists who pioneered 
urban studies pointed out long ago, segregation sharpens the boundaries 
of residential areas, whereas transportation tends to blur them. 1- A good 
example of the destabilizing effects of the increased mobility afforded by 
mechanical transportation are the changes that working-class neighbour¬ 
hoods underwent towards the end of the nineteenth century. These 
neighbourhoods had sharply defined borders when the journey to work 
was on foot, but as the electric trolley became available the need to live 
near the factory was removed and new working-class suburbs with more 
porous boundaries emerged. Vance summarizes the situation thus: 

The fundamental assemblage of buildings and uses in the English 
industrial city was the working class district composed of row housing 
ranged around one or several factories and served by quite local shops 


and pubs. The locating factor was the factory, because the hours of* 
labour were long and the virtually universal way of going to work was! 
on foot. The result was the creation of a city, or even a metropolis, off 
small, very definite neighbourhoods, which contained the life of most 
people save for weekly or less frequent visits to the market square, the 
market hall, or the street market for the buying of items of clothing, 
house furnishings, or perishable food. This parochial existence was 
enforced by conditions of work and housing and the economic { 
unavailability of access to mechanical transportation. Only later in the 
nineteenth century, when the bicycle, the trolley, and finally the ■ 
cheap excursion to the seaside by train began to come into the life of 
the working class, did any appreciable breaking out of this narrow 
geographical frame of life occur. 18 

Increased geographical mobility, in turn, interacted with the way in 
which land-assignment and land-use were determined to produce more j 
drastic changes in the identity of neighbourhoods. Central authorities ■ 
have always had a say in these allocative decisions, and they still do, their : 
zoning regulations having a territorializing effect. Land-rents, on the ; 
other hand, when they became sufficiently fluid to give rise to economic f 
speculation, were a powerful deterritorializing force, divorcing the t 
reasons for land-ownership from any consideration of the activities 
taking place in it and promoting the relatively rapid displacement of one 
land-use by another. Early urban sociologists referred to this phenom- 
enon as land-succession, after the ecological process in which a given ; 
assemblage of plants gives way to another assemblage as an ecosystem 
grows towards its climax mix of vegetation. Instead of plants these 
sociologists were concerned with land-uses and modelled this succession 
as a concentric expansion away from a city's centre. The core was taken 
over by a central business district, encircled by a zone in transition, with 
light manufacture and deteriorating residential neighbourhoods. Next 
came a ring of working-class neighbourhoods, followed by middle- and 
upper-class neighbourhoods, and finally the suburbs or the commuters' 
zone. 1 ’ 

Those early studies, however, focused on a single city (Chicago) and 
did not give a full explanation of the mechanisms involved in succession. 
The concentric-ring model seems to be valid for many cities in the USA, 
where incomes do tend to rise with distance from a city's centre, but not 
for many parts of Continental Europe, where the reverse is the case. 20 



This may explained by the older age of European cities and the fact that, 
as I mentioned before, proximity to the centre was very prestigious earlier 
in their history. At the core, the displacement of residential by 
commercial uses in the nineteenth century was a kind of territorial 
invasion which produced the central shopping district. While a whole¬ 
saler's location was determined by proximity to the port or the railroad 
station, the location of retail shops became increasingly determined by 
the intensity of pedestrian traffic and the convergence of transportation 
lines. 21 Having conquered its territory near the centre, retail itself 
differentiated into specialty shops (with more locational freedom) and 
commodity-combining shops, such as the centrally located department 
store, the first example of which emerged in Paris in the 1850s. 22 In 
addition, retailing had to compete with activities involving the exchange 
of information - as it occurs among brokers, bankers, couriers and other 
traffickers of knowledge - and its shops with the office space sought out 
by these service providers. Eventually, taller buildings decreased the 
intensity of the competition by giving the territory a vertical differentia¬ 
tion, with shops occupying the first floor and offices those higher up. 

Explaining the process of land-succession already involves going 
beyond individual neighbourhoods to a consideration of populations or 
collectivities of neighbourhoods interacting with one another. Moreover, 
since these interactions depend on the relative location of members of 
these populations with respect to a central locale, land-succession implies 
the existence of larger assemblages of which neighbourhoods and districts 
are component parts: towns and cities. Tire identity of these larger 
assemblages, in turn, may be affected by the succession processes taking 
place within them. As I argued above, the centre of a city, particularly 
when there is a single one, is a privileged locale which plays a large role in 
defining its identity. A central square may owe its location to the building 
which served as a nucleus for the urban settlement, a church or a castle, 
lor example, and to this extent may serve as an expression of the 
historical origins of the town. Likewise, when the centre is occupied by a 
market-place, the commercial character of the town is expressed by that 
very fact. Thus, when a city loses its mono centricity its historical identity 
may be affected. This multiplication of centres occurred in many 
countries after 1945 as suburbanization and the increased use of 
automobiles made the city's core a less promising place for retail 
activities, and as shopping centres in outlying locations became 
increasingly common. 




But even before the proliferation of suburbs and industrial hinter® 
lands, the identity of urban settlements depended on their relationship 
with their surroundings. Until relatively recently this meant th«8 
countryside and its rural villages. A town may emerge within a pre« 
existing rural area, a process referred to as synoedsm, or on the contrary, |(K 
may be planted in an area lacking previous rural inhabitants with urba n# 
life projected outwards on surrounding areas, a process called dioedsm.^m 
But whether it is through a rural implosion or an urban explosion that# 
the difference between town and countryside is established, it is thija 
difference that constitutes them both, a difference in their mix of routine! 
activities and in their density of population. The distinction of routine! 
activities is based on the oldest form of division of labour: that between* 
agricultural activities on the one hand, and those of commerce, industry* 
and formal government on the other. Until the last two centuries, this* 
separation of activities was not abruptly discontinuous: towns kept 
vegetable gardens and raised farm animals within their walls, while rural 
villages engaged in small-scale industry.* 4 The distinction in terms of 
demographic density also varied in sharpness, but it was always there, 
however blurred. Some big villages may have been larger than some 
small towns, but the latter always packed more people into the same 
amount of space. 

The relations between town and countryside may be characterized in 
terms of the resources with which they supply one another. A medieval 
town of 3,000 inhabitants, for example, needed the land of about ten 
villages (or 8.5 kilometres) to generate enough food for its inhabitants. 25 
But those villages, in turn, needed services from the town, from the 
commercial services provided by its market-place to the legal, medical 
financial and educational services supplied by its organizations, as well as 
the military protection afforded by its walls and armies. Yet, despite the 
mutuality of resource dependencies, cities have always tended to dominate 
the countryside because of the cumulative, self-stimulating dynamics that 
characterize them. There are many models of these dynamics, some 
stressing the mutual stimulation between the accumulation of workers in 
a place and the availability of economic investment, private or public, in 
that place; others focusing on the mutual stimulation between different 
economic activities that supply each other with materials and services and 
provide demand for each other's products. In all models, however, 'spatial 
concentration itself creates the favourable economic environment that 
supports further or continued concentration'. 26 These self-stimulating 


, mm * h 


dynamics can make towns grow much faster than their countryside, 
increasing their influence and breaking the symmetry of the resource 

In fact, an assemblage analysis of urban centres must take into account 
not only town and countryside, but also the geographical region they 
both occupy. This region is an important source of components playing a 
material role in the assemblage. The geographical site and situation of a 
given urban settlement provides it with a range of objective opportunities 
and risks, the exploitation and avoidance of which depends on 
interactions between social entities (persons, networks, organizations) 
and physical and chemical ones (rivers, oceans, topsoil, mineral deposits). 
In addition to ecological components there are those making up the 
infrastructure of a city, that is, its physical form and its connectivity. 
While the physical form of some towns may result from a mere 
aggregation of its neighbourhoods, some aspects of its connectivity (those 
related to city wide mechanical transportation) tend to have properties of 
their own, and are capable of affecting the form of the neighbourhoods 
themselves. The best example is perhaps that of locomotives. Their large 
mass made them hard to stop as well as to accelerate again, and this 
demanded the construction of elevated or underground tracks whenever 
they had to intermesh with pedestrian traffic. The same physical 
constraints determined an interval of two or three miles between train 
stops, directly influencing the spatial distribution of the suburbs which 
grew around railroad stations, giving this distribution its characteristic 
headlike shape. 27 

The components playing an expressive role in an urban assemblage 
may also be a mere aggregation of those of its neighbourhoods, or go 
beyond these. Let's take for example the silhouette which the mass of a 
town's residential houses and buildings, as well as the decorated tops of 
its churches and public buildings, cut against the sky. In some cases, this 
skyline is a mere aggregate effect, but the rhythmic repetition of 
architectural motifs - belfries and steeples, minarets, domes and spires, 
even smokestacks, water-towers and furnace cones - and the way these 
motifs play in counterpoint with the surrounding features of the 
landscape, may result in a whole that is more than a simple sum. 28 
Either way, skylines, however humble, greeted for centuries the eyes of 
incoming people at the different approaches to a city, constituting a kind 
•f visual signature of its territorial identity. This was particularly true 
before the blurring of city boundaries by suburbs and industrial 



hinterlands, but cities endowed with large skyscrapers continue to posses* 
this physical expressivity even in these new conditions. In some case® ! 
however, as the architectural historian Spiro Kostoff reminds us, thw 
process through which old and new skylines become territorial signature^ 
involve a variety of visual representations, such as those found in coin® | 
paintings and prints aimed at tourists. 29 i 

The processes that stabilize a city's identity concern both the sharpness 
of its physical borders as well as the routine human practices taking placet 
within those borders, in particular, the form taken by residential practices% 
In ancient Greek towns, for example, a substantial part of the population! 
returned to their rural homes in summer months or in times of economics 
trouble. This custom, in turn, affected the process of congregation that! 
formed neighbourhoods within towns: residents tended to congregate bjf| 
their rural place of origin and maintained their geographic loyalties. 30 InJ 
addition, military threats made the inhabitants of a Greek town disperse! j 
rather than hide behind its walls. This combination of factors resulted inf 
towns that, in a sense, blended with their countrysides and therefore didjjf 
not have a sharply defined identity. The opposite case is exemplified by§ 
medieval European towns, where fortified walls provided not only! 
protection for the rural population during a siege but also a sense oft 
security against undefined outsiders: a sense which, even in the absenc® 
of overt conflict, helped to make citizens into dearly defined insiders. Inf 
addition, the stone walls marked the point beyond which the exclusivity! 
of citizenship and its privileges ended, unlike the Greek case in which! 
citizenship could be held by those who practised a duality of residence)! 
Overall, medieval towns had a much sharper identity as locales. These! 
cities, as Braudel writes, 'were the West's first focus of patriotism - and? | 
the patriotism they inspired was long to be more coherent and much; 
more conscious than the territorial kind, which emerged only slowly in? 
the first states'. 31 f 

The native town in ancient Greece and the walled medieval town 
represent two extreme forms which city boundaries may take. An 
interesting intermediate case was created by the rise of the suburb in the 
nineteenth century, and its proliferation in the twentieth. Whereas at 
first suburbs and industrial hinterlands simply blurred the outer 
boundaries of cities which otherwise retained their centre, and hence 
their old identity, after the Second World War not only the area which 
suburbs occupied but the variety of their land-uses (retail, wholesale, 
manufacturing and office space) multiplied, recreating the complex 



combinations that used to characterize theold central business district. As 
[ noted before, this process created brand new centres in the suburban 
band. In some cases, the urban realms around these centres were so self- 
sufficient that the daily paths of their residents could be contained within 
their limits. 32 Thus, by creating a true multi-centred urban space, 
suburban growth - and the changes in connectivity brought about by the 
automobile and the freeway - acted as a powerful deterritorializing force. 

As usual, an assemblage analysis of singular, individual entities must 
be complemented by a study of the populations formed by those entities. 
An important property of populations of towns and cities is the birth-rate 
of new urban settlements, as well as the rate at which old settlements 
disappear. These determine the overall rate of urbanization of a particular 
geographical region. In the case of Europe, urbanization intensified in the 
eleventh and twelfth centuries, accelerated again in the sixteenth, and 
picked up speed once more in the centuries following the Industrial 
Revolution. Between 1 350 and 1450, and between 1650 and 1750, both 
the human population and the overall rate of urbanization declined. 33 
The first wave of city-building took place against the background of 
feudalism, creating densely occupied areas in which a certain autonomy 
from feudal relations could be achieved - the city's land still belonged to a 
bishop or a prince, but the city as a whole paid the rent - as well as areas 
with lower urban density in which cities could not shed their shackles. 

Higher density affected not only the relations of cities with feudal 
organizations, making them more contractual and less directly tributary, 
but also the intensity of the economic interactions between cities. In the 
period between the years 1000 and 1300, cities in the low-density feudal 
areas (Spain, France, England) did not develop systematic relations 
among themselves, remaining within relatively dosed politico-economic 
domains in which trade relationships were mostly local. In the high- 
density areas (northern Italy, Flanders, the Netherlands, some parts of 
Germany), on the other hand, the regularity of trade was greater, its 
volume higher, and it covered much larger areas. This led to the 
generation of more systematic and enduring relations among urban 
centres creating the conditions for the emergence of larger assemblages: 
hierarchies and networks of cities. Much as the differentiation between a 
city and its surrounding countryside involved breaking the symmetry of 
its resource dependencies through self-stimulating accumulations, other 
cumulative processes - related to differential degrees of autonomy from 
feudal organizations, the relative speed of different forms of transporta- 



tion, differences in volume and intensity of trade - destroyed th 
possibility of a uniformly sized population of towns with symmetri 
resource dependencies. 

In formal models of urban dynamics, assemblages of cities of differed 
sizes emerge from a sequence of symmetry-breaking events, as each town 
confronts centripetal processes, like the capture of population, invest 
ment and other resources, as well as centrifugal ones, like congestion, 
pollution, traffic. At the tipping-point, when one set of forces begins to 
dominate the other, a town may grow explosively or shrink to a small size 
in the shadow of a larger one. 34 In computer simulations the actual 
pattern that emerges is not unique - as if there existed a single optimal' 
pattern to which the urban dynamics always tended - but is, on the' 
contrary, highly sensitive to the actual historical sequence of events. For 
this reason, the emergent pattern of urban centres is like a memory of this 
symmetry-breaking sequence 'fossilized in the spatial structure of the 
system'. 35 

A recurrent emergent pattern in these formal models is one familiar to! 
geographers: a hierarchy of central places. In its original formulation,« 


central-place theory was an attempt to describe the hierarchical relations S 

• || 

among regularly spaced urban centres, with larger ones displaying a- : 
greater degree of service differentiation than smaller ones. In the £ 
hierarchies that emerged in medieval Europe, for example, the smallest f 
towns offered a small market-place and a church as services to their rural | 
surroundings; medium-sized towns added to this marketing function f 
more elaborate religious services, as well as some simple administrative i 
and educational ones, such as county jails and schools, which they 
offered to their countrysides as well as to lower-ranked towns. Larger f 
towns, in turn, multiplied the variety of marketing, administrative and 
religious services and added new ones, such as the sophisticated 
educational services provided by universities. 36 In short, in a central- 
place hierarchy each rank offers all the services of the immediately lower 
rank and a few more, and these added services create resource 
dependencies across ranks. To these it must be added the economic 
dependencies which trade may create, since larger towns typically offered 
a larger variety of products than smaller ones, as well as political 
dependencies derived from the fact that the largest towns at the top of the 
hierarchy were usually regional or provincial capitals. In addition to 
landlocked central-place hierarchies, trade among the European popula¬ 
tion of towns in the Middle Ages generated extensive networks of maritime 



ports in which cities were not geographically fixed centres but changing 
relays, junctions or outposts. As the urban historians Hohenberg and Lees 

Instead of a hierarchical nesting of similar centres, distinguished 
mainly by the number and rarity of services offered, [a maritime 
network] presents an ordering of functionally complementary cities 
and urban settlements. The key systemic property of a city is nodality 
rather than centrality, whereas hierarchical differences derive only 
partly from size and more from the nature of the dominant urban 
function. Control and innovation confer the most power and status, 
followed by transmission of goods and messages, and finally by 
execution of routine production tasks. Since network cities easily 
exercise control at a distance, the influence of a town has little to do 
with propinquity and even less with formal control over territory. 37 

Each node in these networks specialized on a subset of economic 
activities not shared with the rest, with the dominant nodes typically 
monopolizing those that yielded the most profits. Since rates of profit vary 
historically, as sources of supply change or as fashion switches demand 
from one luxury product to another, the mix of activities in each node of 
the network also changed, and this, in turn, affected the dominance 
relations between nodes. For this reason, the position of dominant node, 
or 'core', as it is sometimes referred to. changed over time, although it 
was always occupied by a powerful maritime port. The sequence of cities 
occupying the core was roughly this: Venice was dominant in the 
fourteenth century, followed by Antwerp in the fifteenth, Genoa in the 
sixteenth, Amsterdam in the seventeenth, London in the next two 
centuries, and New York in the twentieth. 38 Besides economic specializa¬ 
tion, Hohenberg and Lees mention control at a distance as a characteristic 
of city networks, a relative independence from spatial proximity made 
possible by the much higher speed of transportation by sea relative to that 
over land. Faster transportation implied that nodes in the network were 
in a sense closer to each other than to the landlocked cities in their own 
backyard: news, goods, money, people, even contagious deceases, all 
travelled more rapidly from node to node than they did from one central 
place to another. 

As assemblages, central-place hierarchies and maritime networks have 
different components playing material and expressive roles. Materially, 



they vary in both geographical situation and connectivity. On the oi^R 
hand, the geographical siting of central places always gave them commanM 
over land resources, farmland in particular. By contrast, the cities j2 
maritime networks, particularly the dominant nodes, were relatively poriB 
in these terms: Venice was so ecologically deprived it was condemned tM 
trade from the start, and Amsterdam had to be constantly reclaiming larJp 
from the sea. In terms of connectivity, roads linked central place# 
following the ranks of the hierarchy: there were seldom direct land-rout^® 
connecting the smaller towns to the regional capital. Also, the relativtjj 
slowness of terrestrial transportation forced towns to cluster together* 
since the services offered by larger centres could only be enjoyed if thet, 
smaller ones were located at relatively short distances: the distance it# 
inhabitants would be willing to walk to get the needed service. Maritime! 
ports were not subject to these constraints. Not only were long distance# 
less of a problem, given the faster speed of their ships, but they could all bet 
directly connected to one another regardless of rank. The key to thisf 
connectivity was the sea. Buring the first wave of urbanization, foffi 
instance, 'the two inland seas, the Mediterranean-Adriatic and th^| 
Channel-North Sea-Baltic, served to unite trading centres rather than# 
to separate them'. J9 After that, first the Atlantic Ocean, and later on th<J 
Pacific, became the connecting waters of a network that by th®| 
seventeenth century had acquired global proportions. f 

While the expressive components of these assemblages may be a merof 
aggregate of those of the towns that are their component parts, I he ’• 
aggregate may have a pattern of its own. In the case of central places, if we f 
imagine travelling from the smallest and simplest towns up the ranks until * 
we reach the regional capital, this experience would reveal a pattern off 
increased complexity in the expressive elements giving towns their| 
personality: taller and more decorated churches and central plazas, more 
lavish religious and secular ceremonies, a greater variety of street and 
workshop activities, as well as more diversified and colourful market¬ 
places. In the case of maritime networks, it was not the increased 
differentiation of one and the same regional culture that expressed a 
dominant position but the gathering of expressions from all over the world. 
The core cities, in particular, always had the highest cost of living and the 
highest rate of inflation, so every commodity from around the world, 
however exotic, tended to flow towards their high prices. ‘These world- 
cities put all their delights on display', writes Braudel, becoming universal 
warehouses, inventories of the possible, veritable Noah's Arks. 4 * 


Territorialization in these assemblages is performed by the processes 
that give an entire region a certain homogeneity. The largest central 
places, often playing the role of political capitals, attracted talented people 
from the lower-ranked towns: people who brought with them linguistic 
and nonlinguistic elements of their own local culture. Over time, these 
capitals gathered, elaborated and synthesized these elements into a more 
or less homogenous product which was then re-exported back to the 
smaller centres. 41 The higher prestige of the more differentiated culture at 
the top acted as a magnet for the short-distance migratory patterns of 
cultural producers, and gave the synthesized cultural product the means 
to propagate throughout the region. Long-distance trade, on the other 
hand, had deterritorializing effects. The nodes of a maritime network 
often played the role of gateways to the outside, opening up to foreign 
civilizations, so they housed a more colourful and varied population. 
Having a larger proportion of foreign merchants than did the central 
places, maritime ports offered their inhabitants the opportunity to be in 
more regular contact with outsiders and their alien manners, dress and 
ideas. The existence of dominant nodes implies that the more 
cosmopolitan culture of urban networks was not egalitarian, but its 
heterogeneity was preserved since it was 'superimposed on a traditional 
periphery with no attempt at integration or gradual synthesis'. 42 

Moving from the scale of city assemblages to that of territorial states 
may be done in an abstract way, simply noting that the landlocked 
regions organized by central-place hierarchies and the coastal regions 
structured by maritime networks are today component parts of nation¬ 
states. But this would leave out the historical process behind the 
absorption of cities into larger entities, as well as the resistance offered by 
urban centres to such an integration. In Europe, the outcome of this 
process varied, depending on the segment of the population of cities that 
was involved. In the densely urbanized regions cities managed to slow 
down the crystallization of territorial states until the nineteenth century, 
while in the areas of low density they were quickly absorbed. In 
particular, unlike the central-place hierarchies just examined, those that 
emerged in the areas where feudalism remained dominant tended to 
adopt distorted forms with excessively large cities at the top. These 
disproportionately populous and powerful centres formed the nucleus 
around which empires, kingdoms and nation-states grew by a slow 
accretion of territory, and, in time, they became the national capitals of 
these larger assemblages. 




Although the incorporation of cities in the sixteenth and seventeent 
centuries was performed through a variety of means, direct miiitai 
interventions were often involved. In some cases the rulers of kingdom 
or empires made claims to the territory on which cities were located 
claims legitimated by inheritance or marriage but often enforced throng 
the use of organized violence. But warfare also influenced the outcome 
the contest between cities and territorial states indirectly through tt„ 
enormous expense that armies and fortified frontiers implied. Only large, 
centralized governments, commanding the entire resources of a land ant 
its inhabitants, could afford to stay in the arms races that developed 
between new weapons (such as mobile artillery) and defensive fortifica-Jt 
tions. As the historian Paul Kennedy writes; j* 

Military factors - or better, geostrategical factors - helped to shape the* 
territorial boundaries of these new nation-states, while the frequent! 
wars induced national consciousness, in a negative fashion at least, inf 
that Englishman learned to hate Spaniards, Swedes to hate Danesrf 
Dutch rebels to hate their former Habsburg overlords. Above all, it was! 
war - and especially the new techniques which favoured the growth! 
of infantry armies and expensive fortifications and fleets - which! 
impelled belligerent states to spend more money than ever before, and .' 
to seek a corresponding amount in revenues ... In the last few years of * 
Elizabeth's England, or in Phillip IPs Spain, as much as three-quarters : 
of all government expenditures was devoted to war or to debt " 
repayments for previous wars. Military and naval endeavors may not 
always have been the raison d'etre of the new nation-states, but it; 
certainly was their most expensive and pressing activity. 43 

The historical period that sealed the fate of autonomous cities can be 
framed by two critical dates, 1494 and 1648, a period that witnessed 
warfare increasing enormously in both intensity and geographical scope. 
The first date marks the year when the Italian city-states were first 
invaded and brought to their knees by armies from beyond the Alps: the 
French armies under Charles VIII whose goal was to enforce territorial 
claims on the kingdom of Naples. The second date celebrates the signing 
of the peace treaty of Westphalia, ending the Thirty Years War between 
the largest territorial entity at the time, the Catholic Habsburg empire, 
and an alliance between France, Sweden and a host of Protestant-aligned 
states. When the peace treaty was finally signed by the exhausted 


participants, a unified, geopolitically stabilizing Germany had been 
created at the centre of Europe, and the frontiers that defined the 
identity of territorial states, as well as the balance of power between 
them, were consolidated. Although the crucial legal concept of 
-sovereignty' had been formalized prior to the war (by Jean Bodin in 
1576) it was during the peace conference that it was first used in practice 
to define the identity of territorial states as legal entities. 44 Thus, 
international law may be said to have been the offspring of that war. 

As 1 argued in the previous chapter, it is important not to confuse 
territorial states as geopolitical entities with the organizational hierarchies 
that govern them. Geopolitical factors are properties of the former but not 
of the latter. As Paul Kennedy argues, given the fact that after 1648 
warfare typically involved many national actors, geography affected the 
late of a nation not merely through 

such elements as a country's climate, raw materials, fertility of 
agriculture, and access to trade routes - important though they all 
were to its overall prosperity - but rather [via] the critical issue of 
strategical location during these multilateral wars. Was a particular 
nation able to concentrate its energies upon one front, or did it have to 
fight on several? Did it share common borders with weak states, or 
powerful ones? Was it chiefly a land power, a sea power, or a hybrid - 
and what advantages and disadvantages did that bring? Could it easily 
pull out of a great war in Central Europe if it wished to? Could it 
secure additional resources from overseas? 45 

But if territorial states cannot be reduced to their civilian and military 
organizations, the latter do form the main actors whose routine activities 
give these largest of regionalized locales their temporal structure. A good 
example of the new organizational activities that were required after 
1648 were the fiscal and monetary policies, as well as the overall system 
of public finance, needed t« conduct large-scale warfare. On the 
economic side there were activities guided by a heterogeneous body of 
pragmatic beliefs referred to as 'mercantilism'. The central belief of this 
doctrine was that the wealth of a nation was based on the amount of 
precious metals (gold and silver) that accumulated within its borders. This 
monetary policy, it is clear today, is based on mistaken beliefs about the 
causal relations between economic factors. On the other hand, since one 
means of preventing the outward flow of precious metals was to 




discourage imports, and this, in turn, involved the promotion of locajf 
manufacture and of internal economic growth, mercantilism ha<# 
collective unintended consequences that did benefit territorial states 
the long run. 46 For this reason, however, it is hard to consider the peopu| 
making mercantilist policy decisions the relevant social actors in this case* 
Another reason to consider the activities of organizations the main sourc* 
of temporal structure for territorial states is that many of the capacities^ 
necessary to conduct a sound fiscal policy were the product of 
organizational learning , a feat first achieved in England between the year^f 
of 1688 and 1756. As Braudel writes: 3j 


This financial revolution which culminated in a transformation of f 
public credit was only made possible by a previous thoroughgoing - 
remodeling of the kingdom's finances along clearly defined line* 
Generally speaking, in 1640 and still in 1660, English financial 
structures were very similar to those of France. On neither side of the 
Channel did centralized public finance, under the exclusive control of 
the state, exist. Too much had been abandoned to the private initiative 
of tax-collectors, who were at the same time official money lenders, to 
financiers who had their own affairs in mind, and to officeholders who 
did not depend on the state since they had purchased their posts, not 
to mention the constant appeals that were made to the City of London, 
just as the king of France was always calling on the goodwill of Paris. 
The English reform, which consisted in getting rid of parasitic 
intermediaries, was accomplished steadily and with discretion, though 
without any discernible plan of action 47 

An assemblage analysis of organizational hierarchies has already been ; 
sketched in the previous chapter, so what remains to be analysed is the 
territorial states themselves. Among the components playing a material 
role we must list all the resources contained within a country's frontiers, 
not only its natural resources (agricultural land and mineral deposits of 
coal, oil, precious metals) but also its demographic ones, that is, its 
human populations viewed as reservoirs of army and navy recruits as 
well as of potential taxpayers. As with all locales, the material aspect also 
involves questions of connectivity between regions: questions that in this 
case involve the geographical regions previously organized by cities. 1 
Territorial states did not create these regions, nor the provinces that 
several such regions formed, but they did affect their interconnection 



through the building of new roads and canals. This is how, for example, 
Britain stitched together several provincial markets to create the first 
national market in the eighteenth century, a process in which its national 
capital played a key centralizing role. And, as Braudel argues, without the 
national market 'the modern state would be a pure fiction'. 4 * 

Other countries (France, Germany, the USA) accomplished this feat in 
the following century through the use of locomotives and telegraphs. The 
advent of steam endowed land transportation with the speed it had 
lacked for so long, changing the balance of power between landlocked 
and coastal regions and their cities, and giving national capitals a 
dominant position. With the rise of railroads, as Hohenberg and Lees 
write, although 

many traditional nodes and gateways continued to flourish, the pull of 
territorial capitals on trade, finance, and enterprise could grow 
unchecked. With their concentration of power and wealth, these 
cities commanded the design of rail networks and later of the 
motorways, and so secured the links on which future nodality 
depended. Where once the trade routes and waterways had 
determined urban locations and roles in the urban network, rail 
transportation now accommodated the expansion needs of the great 
cities for both local traffic and distant connections. 49 

On the expressive side, the most important example was the use of 
national capitals as a means to display central control. This was achieved 
through the so-called 'Grand Manner' of urban design pioneered in 
Europe by the absolutist governments of the seventeenth and eighteenth 
centuries. Italian cities created the basic elements of the Grand Manner, 
but it was in France after 1650 that these elements became codified into a 
style: residential blocks with uniform facades acting as frames for 
sweeping vistas which culminated with an obelisk, triumphal arch, or 
statue, acting as a visual marker; long and wide tree-lined avenues; a use 
of the existing or modified topography for dramatic effect; and the 
coordination of all these elements into grand geometric configurations. 5 * 
Although the use of symbols and visual representations was also part of 
this global approach to urban design, it can be argued that the overall 
theatricality of the Grand Manner, and its carefully planned manipulation 
of a city's visual experience, physically expressed the concentration of 
power. To quote Spiro Kostoff: 



If the Grand Manner is routinely associated with centralized pow« 
we can readily see why. The very expansiveness it calls for, and tM 
abstraction of its patterns, presuppose an untangled decision-makin 
process and the wherewithal to accomplish what has been laid 0 \n 
When such clearcut authority cannot be had the Grand Marmi 
remains on paper ... It was not an accident that Washington was tb 
only American city to celebrate the Grand Manner unequivocally , j 
This was the only city in the United States that had a centralize 
administration, however deputized, being under direct authority 
Congress. Elsewhere one could only resort to persuasion, and try ti| 
advance whatever fragments of the overall plan one could through thi 
tangles of the democratic process ... The presumption of absoluMj 
power explains the appeal of the Grand Manner for the totalitarian! 
regimes of the Thirties - for the likes of Mussolini, Hitler and Stalin, 5 ® 

The stability of the identity of territorial states depends in part on th#* 
degree of uniformity (ethnic, religious, linguistic, monetary, legal) that it*: 
organizations and cities manage to create within its borders. A goojf 
example of homogenization at this scale is the creation of standard! 
languages. In the areas which had been latinized during the Romani 
Empire, for example, each central place hierarchy had its own dominan® 
dialect, the product of the divergent evolution that spoken or vulgar Latin! 
underwent after the imperial fall. Before the rise of national capitals thcl 
entire range of romance dialects that resulted from this divergent! 
differentiation coexisted, even as some cities accumulated more prestigef 
for their own versions. But as territorial states began to consolidate theif? 
grip, the balance of power changed. In some cases, special organizations! 
(official language academies) were created to codify the dialects of the? 
dominant capitals and to publish official dictionaries, grammars and; 
books of correct pronunciation. This codification, however, did not* 
manage to propagate the new artificial languages throughout the entire i 
territory. That process had to wait until the nineteenth century for the 
creation of a nationwide system of compulsory elementary education in 
the standard. Even then, many regions and their cities resisted this 
imposition and managed to preserve their own linguistic identity, a 
resistance that was a source of centripetal forces. Although in some 
countries, such as Switzerland, political stability coexists with multi¬ 
lingualism, in others (Canada, Belgium) even bilingualism has proved to 
be a destabilizing force. 52 



In addition to internal uniformity, territorialization at this scale has a 
inore direct spatial meaning: the stability of the defining frontiers of a 
country. This stability has two aspects, the control of the different flows 
jnoving across the border, and the endurance of the frontiers themselves. 
The latter refers to the fact that the annexation (or secession) of a large 
piece of land changes the geographical identity of a territorial state. 
Although these events need not involve warfare aimed at territorial 
expansion (or civil war aimed at secession) they often do, and this shows 
the importance of deploying armies near the border or constructing 
special fortifications for the consolidation of frontiers. A few decades after 
the treaty of Westphalia was signed, for example, France redirected 
enormous resources to the creation of coherent, defensible boundaries, 
through the systematic construction of fortress towns, perimeter walls 
and citadels - separate star-shaped strongholds sited next to a town's 
perimeter. In the hands of Sebastien le Prestre de Vauban, the brilliant 
military engineer, France's defining borders became nearly impregnable, 
maintaining their defensive value until the French Revolution. Vauban 
built double rows of fortresses in the northern and southeastern frontiers, 
so systematically related to each other that one 'would be within earshot 
of French fortress guns all the way from the Swiss border to the 
Channel'. 53 

Migration and trade across national borders tend to complicate the 
effort to create a single national identity, and to this extent they may be 
considered deterritorializing. The ability to reduce the permeability of 
frontiers depends to a large degree on the conditions under which a 
territorial entity comes into being. Those kingdoms and empires that 
crystallized in the feudal areas of Europe had an easier task creating 
internal homogeneity than those in the densely urbanized areas that had 
to cope with the split sovereignty derived from the coexistence of many 
autonomous city-states. 54 Similarly, territorial states born from the 
collapse of a previous empire or from the break-up of former colonial 
possessions can find themselves with unstable frontiers cutting across 
areas heterogeneous in language, ethnicity or religion: a situation which 
militates against a stable identity and complicates border control. A more 
systematic challenge to border control and territorial stability has existed 
since at least the seventeenth century. As the ide tity of the modern 
international system was crystallizing during the Thirty Years War, the 
city of Amsterdam had become the dominant centre of a transnational 
trade and credit network that was almost as global as anything that exists 




today. If the rise of kingdoms, empires and nation-states exerted 
territorializing pressures on cities by reducing their autonomy, maritime,> 
networks not only resisted these pressures but were capable then, and- 
still are today, of deterritorializing the constitutive boundaries of ! 
territorial states. The pressure on these boundaries has intensified in« ; 
recent decades as the ease with which financial resources can flow across! 
state boundaries, the degree of differentiation of the international ' 
division of labour, and the mobility of legal and illegal workers, have-! 
all increased. y 

That networks of cities, and the transnational organizations based ont 
those cities, can operate over, and give coherence to, large geographical !- 
areas cutting across state boundaries, has been recognized since the?! 
pioneering work of Fernand Braudel, who refers to these areas as 'world- 
economies'. 55 It is too early, however, to tell whether these world- ■ 
economies are as real as the other regionalized locales that have been ; 
analysed in this chapter. Some of the processes that are supposed to | 
endow these economic locales with coherence, such as the synchronized? 
movement of prices across large geographical areas following long § 
temporal rhythms (the so-called 'Kondratieff waves'), remain controver-1 
sial. But what is clear even at this stage of our understanding is that ? 
approaches based on reductionist social ontologies do not do justice to the j 
historical data. This is particularly true of macro-reductionist approaches, | 
such as the so-called 'world-systems analysis' pioneered by Immanuel 
Wallerstein, in which Braudel's original idea is combined with theories of t 
uneven exchange developed by Latin American theorists. 56 In Waller- f 
stein's view, for example, only one valid unit of social analysis has existed 
since the end of the Thirty Years War, the entire 'world-system'. 
Explanations at the level of nation-states are viewed as illegitimate since 
the position of countries in the world-system determines their very 
nature. 57 An assemblage approach, on the other hand, is more 
compatible with Braudel's original idea. Although he does not use the 
concept of 'assemblage', he views social wholes as 'sets of sets', giving 
each differently scaled entity its own relative autonomy without fusing it 
with the others into a seamless whole. 58 

It has been the purpose of this book to argue the merits of such a 
nonreductionist approach, an approach in which every social entity is 
shown to emerge from the interactions among entities operating at a 
smaller scale. The fact that the emergent wholes react back on their j- 
components to constrain them and enable them does not result in a f 



seamless totality. F.ach level of scale retains a relative autonomy and can 
therefore be a legitimate unit of analysis. Preserving the ontological 
independence of each scale not only blocks attempts at micro-reductionism 
(as in neoclassical economics) and macro-reductionism (as in world- 
systems analysis) but also allows the integration of the valuable insights 
that different social scientists have developed while working at a specific 
spatiotemporal scale, from the extremely short duration of the small 
entities studied by F.rving Goffman to the extremely long duration of the 
large entities studied by Fernand Braudel. Assemblage theory supplies the 
framework where the voices of these two authors, and of the many others 
whose work has influenced this book, can come together to form a chorus 
that does not harmonize its different components but interlocks them 
while respecting their heterogeneity. 





1. Ian Hacking, The Social Construction of What? (Cambridge, MA: Harvard 
University Press, 1999), p. 103. 

2. Ibid., p. 49. 

3. For passages on assemblage theory, see Gilles Deleuze and Felix Guattari, A 
Thousand Plateaus (Minneapolis, MN: University of Minnesota Press, 1987), 
pp. 7 1, 88-91, 323-37, 503-5. 

4. Manuel DeLanda, Intensive Science and Virtual Philosophy (London: Con¬ 
tinuum, 2002). 

5. Margaret Archer, Realist Social Theory. The Morphogenetic Approach (Cam¬ 
bridge: Cambridge University Press, 1995). Archer does a simi lar critique of 
sociological theories but speaks of 'conflation' rather than 'reduction'. My 
micro-reductionism, macro-reductionism and meso-reductionism are la¬ 
belled downward conflation', ‘upward conflation' and 'central conflation' 
by her. 

6. Manuel DeLanda, War in the Age of Intelligent Machines (New York: Zone 
Books, 1991); Manuel DeLanda, A Thousand Years of Non-Linear Histoty (New 
York: Zone Books, 1997). 

Chapter 1 

1. Howard Becker and Harry timer Barnes, Social Thought from Lore to Science 
(New York: Dover, 1961). pp. 677-8. 

2. G.W.F. Hegel, The Science of Logic (Amherst, NY: Humanity Books, 1999), 
Volume 2, Book 2. p. 711, (Emphasis in the original), 

3. 'Structure is not "external'' to individuals: as memory traces, and as 

instantiated in social practices, it is in a certain sense more "internal" than 
exterior to their activities in a Durkheimian sense' (Anthony Giddens, The 
Constitution of Society [Berkeley, CA: University of California Press, 1986). p. 

4. Ibid., page 3. 

5 . Anthony Giddens, Central Problems in Social Theory (Berkeley, CA: University 
of California Press, 1979), p. 53. 

6. Mario Bunge, Causality and Modem Science (New York: Dover, 1979), p. 156, 

7. Gilles Deleuze and Claire Parnet, Dialogues II (New York: Columbia 
University Press, 2002), p. 55. 

8. Gilles Deleuze, Empiricism and Subjectivity (New York: Columbia University 
Press, 1991), p. 98. Deleuze is here discussing a specific type of component, 
Humean ideas (and this is what the original quote refers to), but the point 
applies to any other type of component. 

9. Thus Deleuze writes: 

What is an assemblage? It is a multiplicity which is made up of 
heterogeneous terms and which establishes liaisons, relations between 
them, across ages, sexes and reigns - different natures. Thus the 
assemblage's only unity is that of a co-functioning: it is a symbiosis, a 
'sympathy'. It is never filiations which are important, but alliances, 
alloys; these are not successions, lines of descent, but contagions, 
epidemics, the wind. (Deleuze and Parnet, Dialogues II, p. 69) 

The exclusion of lines of descent, such as they exist among organisms and 
even species, shows that he means to exclude the latter from the definition of 
an assemblage. In his work with Felix Guattari, Deleuze distinguishes between 
'assemblages' on the one hand, and 'strata' on the other. Biological organisms 
and institutional organizations would be classified by them as strata. I will not 
retain this distinction here for reasons explained below in note 21. 

10. Deleuze and Guattari use slightly different terminology. In particular, 
instead of 'material' and 'expressive' roles for components they talk of 
segments of ‘content' and 'expression': 

We may draw some conclusions of the nature of Assemblages from this. 
On a first, horizontal axis, an assemblage comprises two segments, one of 
content, the other of expression. On the one hand it is a machinic 
assemblage of bodies, of actions and passions, and intermingling of bodies 
reacting to one another; on the other hand, it is a colleaivc assemblage of 
enunciation, of acts and statements, of incorporeal transformations 
attributed to bodies. Then, on a vertical axis, the assemblage has both 
territorial sides, or reterritorialized sides, which stabilize it, and cutting 
edges of deterritorialization, which carry it away. (Gilles Deleuze and Felix 
Guat tari, A Thousand Plateaus [Minneapolis, MN: University ofMinnesota 
Press, 1987], p. 88) 




With the exception of the term 'territorialization' I will avoid using any gde 
this complex terminology in this book. Also, instead of two dimensions I use® 
three, a manoeuvre which allows me to get rid of the distinction between 
strata and assemblages, as explained in note 21. 

11 This distinction between linguistic and nonlinguistic expression i s somewhat 
obscured in the previous note by the reference to expressive components as- 
'collective assemblages of enunciation', unless one interprets it as referring 
not to the semantic content of statements, but to their illocutionary force, 
that is, to what they express as 'speech acts'. See Deleuze and Guattari, A 
Thousand Plateaus, p. 80. 

At any rate, even if we interpret 'statement' this way, the definition of 
assemblage is still inconvenient in that it seems to apply only to social cases 
(unless one takes inorganic and biological entities as capable of producing 
statements) which goes directly against the idea that assemblage theory 
applies equally well to physics, biology and sociology. See also note 13. 

12. Edwin C. Kemble, Physical Science. Its Structure and tevetopment (Cambridge, 
MA: MIT Press, 1966), pp. 126-7. 

13. Deleuze and Guattari, A Thousand Plateaus, p. 62. Deleuze and Guattari 
distinguish the substance and the form of the materiality and expressivity of 
assemblages. Materiality involves not merely substance but formed substance, 
and expressivity is not purely formal but it involves its own substance. The 
specialization of genes and words is then conceptualized as the separation 
between the substance and form of expression. In what follows I will not stick 
to this terminology. I will speak of physical or direct expressivity to refer to, 
for example, facial expressions or the expressivity of behaviour, and refer to 
language as a specialized medium of expression. But the reader should keep in 
mind that facial expressions are referred to by Deleuze and Guattari as 
'substance of expression' and language as 'form of expression'. As they write: 
'On the other hand, language becomes the new form of expression ... The 
substance involved is fundamentally vocal substance, which brings into play 
various organic elements: not only the larynx, but the mouth and lips, and 
the overall motricity of the face' (ibid., p. 61). 

14. In addition, the processes which territorialize or deterritorialize genes and 
words should be included. The materiality of language, for example, 
becomes territorialized with the emergence of writing. But this spatial 
identity may become deterritorialized when carvings in stone or inked 
inscriptions on paper become modulations in electromagnetic fields, as in 
radio transmissions of spoken language, or television broadcasts of written 
language. Deteriitorializations of the expressive part of language, that is, its 
semantic content, are trickier to conceptualize. Deleuze gives some 
indications of how this conceptualization may be pursued. In particular, 
he singles out certain semantic entities as playing a key role in these 



processes: infinitive verbs, proper nouns, indefinite articles. See ibid., pp. 

15 . Deleuze and Guattari refer to this synthesis of wholes out of components as 
a process of double articulation (ibid , pp. 40-41). (This process is said to 
synthesize strata not assemblages, but see below, note 21.) 

16. Ibid., p. 316. 

17. Historically, the ancient Greek cities, located far from their main 
contemporary empires, but not so far that they could not benefit from their 
advanced civilizations, may have supplied the conditions in which 
conversations between friends broke free from the rigidity of similar 
encounters elsewhere. See Gilles Deleuze and Felix Guattari, What is 
Philosophy ? (New York: Columbia University Press, 1994), p. 87. The Greek 
case is in fact a combination of deterritorialization and decoding. Here 
Deleuze and Guattari stress the former, but I would argue that decoding is 
also involved. 

18. Fernand Braudel, The Perspective of the World (New York: Harper & Row, 
1979), pp. 280-82. 

19. Ibid., pp. 282-4. 

20. Ibid., p. 287. 

21. This departs from Deleuze and Guattari's own version of assemblage theory 
since they define assemblages along two, nor three dimensions, but they are 
then forced to introduce two categories of actual entities, strata and 
assemblages. To use this opposition would unnecessarily complicate the 
presentation, particularly when the same objective may be achieved by 
adding a third dimension to the concept of assemblage. That they thought 
the opposition between strata and assemblages was relative (i.e. that 
assemblages are a kind of strata, or vice versa) is clear from the following: 

From this standpoint, we may oppose the consistency of assemblages to 
the stratification of milieus. But once again, this opposition is only 
relative, entirely relative. Just as milieus swing between a stratum state 
and a movement of destratification, assemblages swing between a 
territorial closure that tends to restratify them and a deterritorializing 
movement that connects them to the Cosmos. Thus it is not surprising 
that the distinction we were seeking is not between assemblage and 
something else, but between two limits of any possible assemblage. 
(Deleuze and Guattari, A Thousand Plateaus, p. 337) 

In addition, Deleuze distinguishes between two forms of deterritorializa¬ 
tion. The first form, relative deterritorialization, refers to processes which 
destabilize the identity of an assemblage, opening it up to transformations 
which may yield another identity (in a process called 'reterritorialization'). 
The second form is quite different, and it is referred to as absolute 
deterritorialization. In this second form it involves a much more radical 




identity change: indeed, a loss of identity altogether, but without falling into 
an undifferentiated chaos. Assemblages exist as actual entities, but the 
structure of the processes of assembly (what gives these processes their 
recurrent nature, or what explains that they can be repeated in the first 
pLace) is not actual but virtual. When deterritorialization is absolute it means 
that the process has departed from actual reality to reach the virtual 
dimension. In this sense, the term is synonymous with 'counter-actualiza¬ 
tion' as the limit process which creates the plane of immanent multiplicities 
which define the virtual structure of assemblages. The two limits referred to 
in the quote above are, on the one hand, a highly territorialized and coded 
assemblage and, on the other, the plane of immanence containing the 
virtual structure of all assemblages linked by relations of exteriority. In 
Chapter 2 I discuss the question of the virtual structure of assemblages using 
the concept of the 'diagram' of an assemblage. 

22. Bunge, Causality and Modern Science, p. 47. 

23. Ibid., p. 175. Bunge credits both Spinoza and Leibniz with the introduction 
of efficient inner causation. Gilles Deleuze continues this tradition when he 
gives equal importance to capacities to affect and capacities to be affected. 

24. Ibid., 49. 

25. Wesley C. Salmon, Scientific Explanation and the Causal Structure of the World 
(Princeton, NJ: Princeton University Press, 1954), pp. 30-34. 

26. Bunge. Causality and Modern Science, pp. 100-1. 

27. R.S. Peters, The Concept of Motivation (London: Routledge & Kegan Paul, 
1960), p. 29. 

25. Max Weber, The Theory of Social and Economic Organization (New York: Free 
Press of Glencoe, 1964), p. 99. 

29. The concept of culture 1 espouse ... is essentially a semiotic one. 
Believing, with Max Weber, that man is an animal suspended in webs of 
significance he himself has spun, I take culture to be those webs, and the 
analysis of it to be therefore not an experimental science in search of law 
but an interpretive one in search of meaning (Clifford Geertz, 'Thick 
description: toward an interpretive theory of culture', in The Interpretation 
of Culture [New York: Basic Books, 1973), p. 5 Imy emphasis]) 

Geertz goes on to speak of 'structures of signification', as if this expression 
meant the same thing as 'webs of significance', a manoeuvre which 
illustrates the error I am discussing here. On the other hand, it must be 
admitted that Geertz's 'thick descriptions' of cultural practices are indeed 
invaluable as a starting point in any social explanation, and this regardless of 
his rejection of explanatory strategies in favour of descriptive ones. 

30. Weber, Theory of Social and Economic Organization, p. 91. 

31. Ibid., p. 116. 

32. Ibid,, p. 115. Weber discusses four ideal types of social action: (1) action 

oriented towards the matching of means to individually chosen ends: (2) 
action oriented emotionally; (3) action oriented by habituation to a 
tradition; and (4) action oriented towards an absolute value, that is, action 
involving a conscious belief in the absolute value of some ethical, aesthetic, 
religious, or other form of behaviour, entirely for its own sake and 
independently of any prospects of external success'. 

33. Ibid., p. 117. 

34. 'Thus causal explanation depends on being able to determine that there is a 
probability, which in the ideal case can be numerically stated, but is always 
in some sense calculable, that a given event (overt or subjective) will be 
followed or accompanied by another event' (ibid., p. 99). 

Chapter 2 

1. Aristotle, The Metaphysics (Buffalo, NY:, Prometheus Books, 1991), p. 155. 

2. One is called that which subsists as such according to accident in one 
way, and in another, that which subsists essentially. A thing is called one 
according to accident, for instance Coriscus and what is musical, and the 
musical Coriscus; for it is one and the same thing to say, Coriscus and 
what is musical, as to say, Coriscus the musician; also, to say the musical 
and the just is one with saying the just musician Coriscus. For ail these 
are called one according to accident, (Ibid., p. 97) 

3. 'The very nature of a thing will not, accordingly, be found in any of those 
things that are not the species of a genus, but in these only, for these seem to 
be predicated not according to participation or passion, nor as an accident' 
(ibid., p. 136). 

4. Michael T. Ghiselin, Metaphysics and the Origin of Species (Albany, NY: State 
University of New York, 1 997), p. 75. 

5. For a full discussion of the ontological and epistemological aspects of phase 
space, see Manuel DeLanda, Intensive Science and Virtual Philosophy (London: 
Continuum, 2002), Ch. 1. 

6 . For Deleuze's most extended discussion of diagrams, see Gilles Deleuze, 
Foucault (Minneapolis, MN: University of Minnesota Press, 1955), pp. 34-41 
and 71-2. 

The structure of a space of possibilities is sometimes referred to as a 
'multiplicity', a term that in French is equivalent to 'manifold', the 
differential geometry spaces used in the construction of phase space. 
Deleuze sometimes uses the terms 'multiplicity' and diagram' as synonyms. 
Thus, he says that 'every diagram is a spatio-temporal multiplicity’ (ibid., 
p. 34). But he also uses alternative formulations that do not involve the 
mathematics of phase space. Thus he defines a diagram as a display of 
relations of force, or of a distribution of capacities t o affect and be affected 




(ibid., pp. 71-2). Since capacities may exist without being exercised (i.e. 
since they may exist as possibilities) they form a possibility space, and a 
diagram would display whatever structure this space has. Elsewhere, his 
definition departs from this spatial form. He argues that unlike an 
assemblage where the material and expressive roles (or the content and 
the expression) are clearly distinguished, the diagram of an assemblage 
involves unformalized functions and unformed matter. This means that diagrams 
have an abstract structure in which the expressive and the material are not 
differentiated, a differentiation that emerges only when the diagram is 
divergently actualized in concrete assemblages. One way of thinking about 
the status of diagrams is, therefore, as the product of a full deterritorializa- 
tion of a concrete assemblage, since it is the opposite process (territorializa- 
tion or actualization) that differentiates the material from the expressive. 
See Gilles Deleuze and Felix Guattari, A Thousand Plateaus (Minneapolis, 
MN: University of Minnesota Press, 1987), p. 142. 

Finally, while 'multiplicity' and 'diagram' are sometimes used inter¬ 
changeably, at other times they refer to separate entities: the structure of a 
possibility space, on the one hand, and the agency responsible for the 
absolute deterritorialization, the abstract machine or quasi-causal operator, 
on the other. For a detailed explanation of these notions and their relations, 
see DeLanda, Intensive Science and Virtual Philosophy. Chs 2 and 3. 

7. Because Deleuze does not subscribe to the multiscale social ontology that I 
am elaborating here, he never says that each of these entities (interpersonal 
networks, institutional organizations, cities, etc.) have their own diagram. 
On the contrary, he asserts that the diagram 'is coextensive with the social 
field' (Deleuze, Foucault p. 34). Deleuze gives as examples of 'social fields' 
contemporary 'disciplinary societies', the 'sovereign societies' that came 
before them, 'primitive societies’, 'feudal societies', etc. (ibid., pp. 34-5). In 
the social ontology I am presenting there is no such thing as 'society as a 
whole' or an overall 'social field', so I am breaking in a rather drastic way 
with Deleuze here. 

This implies that the terms 'micro' and 'macro' as used in this book do not 
correspond to Deleuze's 'molecular' and 'molar'. But some correspondence 
may still be achieved: at every level of scale we may have, on the one hand, 
populations of micro-entities, populations characterized by intensive 
properties such as rates of growth, or the rate at which some components 
propagate within them; and, on the other hand, some of the members of 
these populations may be caught into larger macro-entities, regularized and 
routinized. The entities belonging to the populations could be seen as 
'molecular', while the entities caught in the larger aggregates would be 
molar', particularly if the macro-entity is highly territorialized. These 
remarks soften the differences but do not completely eliminate them. For 


the molecular and the molar see Deleuze and Guattari, A Thousand Plateaus, 
p. 217. 

8. Max Weber, The Theory of Social and Hconomic Organization (New York: Free 
Press of Glencoe, 1964), pp. 328-60. 

9. William Bechtel and Robert C. Richardson, Discovering Complexity. Decom¬ 
position and Localization Strategies in Scientific Thought (Princeton, NJ: 
Princeton University Press, 1993), pp. 52-9. 

10. Gilles Deleuze, Logic of Sense (New York: Columbia University Press, 1990), p. 
169. On the other hand Deleuze sometimes writes about diagrams as if they 
themselves were causes of which assemblages are the effects. Thus he writes 
that 'the diagram acts as a non-unifying immanent cause .., the cause of the 
concrete assemblages that execute its relations' (Deleuze, Foucault, p. 37). 

11. In the last decade the discipline of sociology resuscitated a n old dilemma 
in a new form - a form, unfortunately, that has done little to resolve the 
dilemma itself. The perennial conflict between individualistic and 
collectivistic theories has been reworked as a conflict between micro 
sociology and macro sociology ... I would like to begin by suggesting that 
this equation of micro with individual is extremely misleading, as, 
indeed, is the attempt to find any specific size correlation with the micro¬ 
macro difference. There can be no empirical referents for micro or macro 
as such. They are analytical contrasts, suggesting emergent levels within 
empirical units, not antagonistic empirical units themselves. (Jeffrey C. 
Alexander, Action and its environments', in Jeffrey C. Alexander, 
Bernhard Giesen, Richard Miinch, Neil J. Smelser ]eds], The Micro-Macro 
Link, [Berkeley, CA: University of California Press, 1987], pp. 290-91) 

In the same volume, another sociologist writes: 

A fundamental distinction such as that between micro and macro must 
be general and analytical, not tied to a fixed case. By this standard, the 
individual person, household, or firm cannot be treated as intrinsically 
micro, and the society, nation, or economy as unalterably macro. Rather, 
designations of micro and macro are relative to each other and, in 
particular, to the analytic purpose at hand. The overall status or role of a 
given family member (ego) may be macro relative to ego's relation to a 
certain kin group member, but micro relative to the status or role of ego's 
lineage in a marriage exchange system; the marriage system in turn may 
be micro relative to a mythic cycle. The job satisfaction of a worker may 
be macro relative to the psychological stress on his or her children, but 
micro relative to the quality of his or her job. That in turn may be micro 
relative to the morale or efficiency of the factory or branch office, which 
is micro relative to the financial condition of the corporation, which is 
micro relative to the competitiveness of the industry or the business 
cycle of the national or international economy - which are, however. 



micro relative to the ideological spirit of the age. (Dean R. Gerstein, 'To 
unpack micro and macro: link small with large and part with whole', 
ibid., p. 88) 

12. Roy Bhaskar, A Realist Theory of Science (London: Verso, 1997), p. 114. While 
Bhaskar's realism comes very close to Deleuze's in some aspects it is 
incompatible with it because Bhaskar is a self-declared essentialist. As he 

In general to classify a group of things together in science, to call them by 
the same name, presupposes that they possess a real essence or nature in 
common, though it does not presuppose that the real essence or nature is 
known ... A chemist will classify diamonds, graphite and black carbon 
together because he believes that they possess a real essence in common, 
which may be identified as the atomic (or electronic) structure of carbon. 
(Ibid., p. 210) 

13. Peter Hedstrbm and Richard Swedberg, 'Social mechanisms: an introduc¬ 
tory essay', in Social Mechanisms. An Analytical Approach to Social Theory. 
(eds) Peter Hedstrom and Richard Swedberg (Cambridge: Cambridge 
University Press, 1998), pp. 22-3. The authors propose three different types 
of mechanism: macro-micro, micro-micro and micro-macro. The first type 
would figure in explanations of the relations between a social situation 
involving large sociological phenomena (such as the distribution of income 
or power in a population) and individual social actors. The large-scale 
process may, for example, create different opportunities and risks for 
different actors, who must include these opportunities and risks as part of 
their reasons to act. The second type refers mainly to social-psychological 
mechanisms, that is, to the mental processes explaining the making of 
particular decisions (in the case of motives) or to the processes behind the 
lormation of habits, the production of emotions or the acquisition of beliefs 
(in the case of reasons). Finally, the third type refers to mechanisms 
governing the interactions among individual actors which generate 
collective outcomes. 

The problem is that the terms 'micro' and 'macro' are used in their absolute 
sense, with 'micro' referring to individual persons and 'macro' designating 
society as a whole. But in assemblage theory the distinction between micro- 
and macro-levels is relative to scale. Relativizing the distinction implies that 
their third type of mechanism, micro-micro, can be eliminated since at any 
given scale it reduces to the micro-macro one at the immediately smaller scale. 
And similarly for what we may term macro-macro mechanisms. When 
'macro' refers to 'total society' there is no need to consider the interactions 
between wholes. But once the distinction is relativized we do need to consider 
that wholes made out of individual persons, such as interpersonal networks or 
institutional organizations, may interact with one another as wholes. The 

term macro-macro, however, is not necessary, since it reduces to the micro¬ 
macro case at the immediately larger scale. 

] 4. Mark Granovetter, Getting a Job: A Study of Contacts and Careers (Chicago, 1L: 
University ol Chicago Press, 1995). 

15. David Krackhardt, 'The strength ol strong ties: the importance of philos in 
organizations', in Networks and Organizations, (eds) Nitin Nohria and Robert 
G. Eccles (Boston, MA: Harvard Business School Press, 1992), pp. 218-19. 

1 6. Fernand Braudel, The Wheels of Commerce (New York: Harper & Row, 1979), 
p. 30. 

17. When exactly in the history of Europe prices began to be determined 
impersonally, as opposed to through the decisions of feudal lords, is a 
controversial point. Braudel argues that all 'the evidence relating to prices as 
early as the twelfth century indicates that they were already fluctuating, 
evidence that by then "modern" markets existed and might occasionally be 
linked together in embryonic, town-to-town networks' (ibid., p. 28). 

18. Alan Garfinkel, Forms of Explanation (New Haven, CT: Yale University Press, 
1981), pp. 58-62. 

14. As the sociologist Anthony Giddens argues, unlike the components of a 
physical entity with emergent properties (such as bronze, a metallic alloy 
having properties that are more than the sum of the properties of its pans, 
copper, tin and sometimes lead), the parts of a social assemblage seldom 
come in pure form. It is easy to imagine the component parts of bronze as 
existing separately prior to their coming together and forming an alloy, 'but 
human actors, as recognizable competent agents, do not exist in separation 
from one another as copper, tin, and lead do. They do not come together ex 
nihilo to form a new entity by their fusion or association' (Anthony Giddens, 
The Constitution of Society [Berkeley, CA: University of California Press, 1986], 
pp. 171-2). 

Giddens is th us correct i n criticizing t h e limited concept o f emergence that 
implies only to originary emergence. But he is wrong in thinking that giving 
up this conception implies surrendering the parl-to-whole relation in favour 
of a seamless web. The example of bronze was used by Emile Durkheim to 
argue for the existence of social emergent properties. See Emile Durkheim, 
The Rules of Sociological Method (New York: The Free Press, 1982), p. 39. 

20. Paul DiMaggio, 'Nadel's paradox revisited: relational and cultural aspects of 
organizational structure', in Networks and Organizations, p. 132. 

21. Jeffrey L. Pressman and Aaron Wildavsky, Implementation (Berkeley, CA: 
University of California Press, 1984), p. 92, 

22. This ability to operate across scales is particularly surprising, given that both 
genetic and linguistic materials are 'more micro' than any of the entilies of 
which they form a pari. But Deleuze and Guattari see this 'molecularization' 
ol expression as precisely what gives genes and words their ability to 





produce more complex relations between the micro and the macro. See 
Deleuze and Ciuattari, A Thousand Plateaus, p. 59. 

23. Peter L. Bergerand Thomas Luckmann, The Social Construction of Reality (New 
York: Anchor Books, 1967). 

Chapter 3 

1. 'All the perceptions of the human mind resolve themselves into two kinds, 
which I shall call IMPRESSIONS and IDEAS. The difference betwixt them 
consists in the degrees of force and liveliness with which they strike upon 
the mind, and make their way into our thought and consciousness. Those 
perceptions, which enter with the most force and violence, we may name 
impressions', and under this name I comprehend all our sensations, passions 
and emotions, as they make their first appearance in the soul. By ideas I 
mean the faint images of these in thinking and reasoning ...' (David Hume, 
A Treatise of Human Nature [London: Penguin, 1969], p. 49. [emphasis in the 

2. Ibid., p. 462. 

3. Hume, in fact, makes a distinction between relations which may change 
without changing the related ideas (contiguity, identity, causality) and those 
in which this is not the case (resemblance, contrariety, degrees of quality 
and proportions of quantity) (ibid., pp. 1 17-18). This would seem to 
contradict the statement that all links between ideas are relations of 
exteriority. Yet, as Deleuze argues, this is not so. The four relations which do 
seem to depend on ideas imply a comparison, that is, an operation which is 
exterior to the ideas being compared. See Gilles Deleuze, Empiricism and 
Subjectivity (New York: Columbia University Press. 1991), pp. 99-101. 

4. Hume, A Treatise of Human Nature, p. 60. 

5. As Deleuze puts it: 

... if the principles of association explain that ideas are associated, only 
the principles of the passions can explain that a particular idea, rather 
than another, is associated at a given moment... Everything takes place 
as if the principles of association provided the subject with its necessary 
form, whereas the principles of the passions provided it with its singular 
content. (Deleuze, Empiricism and Subjectivity, pp. 103—1) 

6. Ibid., p. 98. Deleuze is here contrasting an 'assemblage or collection' with a 
'system'. This is similar to the contrast he draws in his latter work between 
assemblages' and 'strata'. As I argued in Chapter 1, I prefer to deal with this 
contrast not as a dichotomy between two types but as a third dimension 
characterizing assemblages, with highly coded assemblages being 'strata'. 

7. Hume, A Treatise of Human Nature, p. 327. 

8. Ibid., p. 51. 

9. Ibid., p. 308. 

10. On the effects of madness see ibid., p. 172. 

1 1. Ibid., p. 308. 

12. The most famous critique of the combinatorial poverty of associationism is 
Jerry A. Fodor and Zenon W. Pylyshyn, 'Connectionism and cognitive 
architecture: a critical analysis', in John Haugeland (ed.), Mind Design 11. 
Philosophy, Psychology and Artificial Intelligence (Cambridge, MA: MIT Press, 
1997), pp. 309-50. 

For a discussion of recent associationist extensions that may compensate 
for this poverty see William Bechtel and Adele Abrahamsen, Connectionism 
and the Mind. An Introduction to Parallel Distributed Processing in Networks 
(Cambridge, MA, and Oxford: Basil Blackwell, 1991), pp. 101-2; Andy 
Clark, Microcognition. Philosophy, Cognitive Science, and Parallel Distributed 
Processing (Cambridge, MA: MIT Press, 1990), pp. 143-51. 

13. A theory of grammar that meets both the combinatorial productivity 
requirement as well as the evolutionary one is Zellig Harris, A Theory of 
Language and Information: A Mathematical Approach (Oxford: Clarendon Press, 
1981). 1 give a fully evolutionary history of real languages and dialects, 
based on Zellig Harris's ideas, in Manuel DeLanda, A Thousand Years of 
Nonlinear History (New York: Zone Books, 1997), Ch. 3. 

14. Hume, A Treatise of Human Nature, p. 144. A belief 'can only bestow on our 
ideas an additional force or vivacity'. 

1 5. Ibid., p. 146. 

16. Ervin Goffman, Interaction Ritual. Essays on Face-to-Face Behaviour (New York: 
Pantheon Books, 1967), p. 1 (my italics). 

17. Ibid., p. 19. 

18. Ibid., p. 103. 

19. Ibid., p. 34. 

20. Ibid., p. 103. 

21. Analytical philosophers, for decades infatuated with syntax and semantics, 
are beginning to turn around and include this pragmatic dimension. Thus, 
Ian Hacking, in his analysis of the term 'social construction', deliberately 
resists asking the question 'what is its semantic content?' and asks instead 
'what is its point?' (i.e. what is its significance?) See Ian Hacking, The Social 
Construction of What? {Cambridge, MA: Harvard University Press, 2000), p. 5. 

An argument that questions of significance are not the same as questions 
of signification can be found in Denis C. Phillips, Philosophy, Science, and Social 
Inquiry (Oxford: Pergamon Press, 1987), p. 109. 

22. Goffman, Interaction Ritual, pp. 162-4. 

23. Ibid., pp. 218-19. 

24. John Scott, Social Network Analysis (London: Sage Publications, 2000), pp. 1 1, 
31 and 75. 





25. Ibid., pp. 70-73. 

26. Ibid., p. 12. 

27. Ibid., p. 79. See also Graham Crow, Social Solidarities (Buckingham: Open 
University Press, 2002), pp. 52-3. 

28. Crow, Social Solidarities, pp. 11 9-20. 

29. On local dialects as badges of identity see William Labov. 'The social setting 
of linguistic change', in Sociolinguistic Patterns (Philadelphia, PN: University of 
Pennsylvania Press, 1972), p. 271. 

30. Crow, Social Solidarities, pp. 1 28-9. 

31. Ibid., pp. 86-8. 

32. Charles Tilly, Stories, Identities, and Political Change (Lanham, MD: Rowman fr 
Littlefield, 2002), pp. 28-9. 

33. Charles Tilly, Durable Inequality (Berkeley, CA: University of California Press, 
1 999), p. 66. 

34. Gilles Deleuze and Felix Guattari, Anti-Oedipus (Minneapolis, MN: University 
of Minnesota Press), pp. 147, 155. 

35. Tilly, Stories, Identities, and Political Change, p. I 2. Tilly is perhaps the most 
coherent advocate of realism in social theory today although his fear of 
essences has made him espouse a rather watered-down version of it. He 
declares himself to be a 'relational realist', that is, someone who believes in 
the mind-independent existence of relations but not of the entities that 
enter into relations, although he grudgingly acknowledges the existence of 
human beings with physiological needs. Enduring entities, in his account, 
presuppose essences and are thus less worthy of commitment. As he puts it, 
social explanations can be either in terms of essences or in terms of bonds. 
See Tilly, Durable Inequality, p. 45. 

But, first of all, a commitment to entities need not involve essences at all if 
the entities are accounted for by the historical process that produced them. 
Secondly, although social interaction is indeed relational, in the sense that 
the capacities exercised by the social actors are not reducible to the actors' 
defining properties, capacities do depend on the existence of these enduring 
properties, and thus, on the existence of enduring entities. Finally, to 
subordinate entities to relations comes dangerously close to a commitment 
to relations of interiority, that is, to wholes in which the pans are 
constituted by the very relations which yield the whole. 

36. Ibid., p. 90. 

37. Ibid., p. 54. 

38. Ibid., p. 89. 

39. Ibid., pp. 106-7. 

40. Ibid., pp. 52-3. 

41. Ibid., pp. 105-6. 

42. Tilly, Stories. Identities, and Political Change. 102-3. 

13 No general population larger than a local community ever maintains a 
coherent system of stratification in a strong sense of the word: even the 
so-called caste systems of India accommodated great variation in rank 
orders from village to village. In general, rank orders remain incon¬ 
sistent, apparent strata contain considerable heterogeneity, and mobility 
blurs dividing lines. (Ibid., pp. 28-9) 

44 . The idea of difference, or gap, is at the basis of the very notion of space, 
that is, a set of distinct and coexisting positions which are exterior to one 
another and which are defined in relation to one another through their 
mutual exteriority and their relations of proximity, vicinity, or distance, as 
well as through relations of order, such as above, below, or between. 
Certain properties of members of the petit-bourgeoise can, for example, 
be deduced from the fact that they occupy an intermediate position 
between two extreme jxtsitions, without being objectively identifiable or 
subjectively identified with one or the other position. (Pierre Bourdieu, 
Practical Reason ((Standford, 1998), CA: Stanford University Press, p. 6 
[emphasis in the original)) 

45. Pierre Bourdieu, The logic of Practice (Cambridge: Polity Press, 1990), p. 54 
(my italics). 

46. Ibid., p. 55. A more generous reading of the habitus, along assemblage 
theory lines, would be as the topological diagram of the set of habits and 
routines that make up individual persons, that is, as the structure of the 
space of possibilities for different combinations of habits and skills. 

47. 'So far as the social world is concerned, the neo-Kantian theory, which gives 
language and, more generally, representations a specifically symbolic 
efficacy in the construction of reality, is perfectly justified. By structuring 
the perception that social agents have of the social world, the act of naming 
helps to establish the structure of this world' (Pierre Bourdieu. language and 
Symbolic Power [Cambridge: Harvard University Press, I991[, p. 105). 

48. Tilly, Durable Inequality, p. 76. 

49. Ibid., p. 36. 

Chapter 4 

). Max Weber, The Theory of Social and Economic Organization (New York: Free 
Press of Glencoe, 1964), p. 331. 

2. Ibid., pp. 328-36. 

3. Ibid,, p. 348. 

4. Ibid., p. 359. 

5. As Weber puts it, even in the most rational bureaucracy the very 'belief in 
legality comes to be established and habitual and this means it is partly 
traditional' (ihid., p. 382). 




6. James S. Coleman, Foundations of Social Theory (Cambridge, MA: Belknap 
Press, 2000), p. 66. 

7. A sick worker must be treated by a doctor using accepted medical 
procedures; whether the worker is treated effectively is less important. A 
bus company must service required routes whether or not there are 
many passengers. A university must maintain appropriate departments 
independent of the departments' enrollment. Activity, that is, has ritual 
significance; it maintains appearances and validates an organization. 
(John W. Meyer and Brian Rowan, 'Institutionalized organizations: 
formal structure as myth and ceremony', in Walter W. Powell and Paul J. 
MiMaggio [eds]. The New Institutionalism in Organizational Analysis 
[Chicago, IL: University of Chicago Press, 1991], p. 55) 

8. W. Richard Scott and John W. Meyer, 'The organization of societal sectors: 
propositions and early evidence', in Powell and DiMaggio (eds). The New 
Institutionalism in Organizational Analysis, p. 124. Valuable as this neo- 
institutional work in sociology may be it is fatally flawed in one sense: it 
relies on social constructivism and its idealist ontology. Hence, despite the 
apparent recognition that there are real technical questions involved in the 
operation of some organizations, ultimately what 'counts as technical' is just 
a mere convention, that is, a matter of definition, an assertion which makes 
the distinction between technical and ceremonial factors useless. 

9. Michel Foucault, Discipline and Punish. The Birth of Prison (New York: Vintage 
Books, 1979), p. 169. 

10. Ibid., p. 171. 

11. Ibid., p. 153. 

12. Ibid., pp. 195-6. 

13. Ibid., pp. 191-2. 

14. Ibid., p. 190. 

15. Weber, The Theoty of Social and Economic Organization, p. 363. 

16. In Deleuze's application of assemblage analysis to Foucault's work, he 
singles out the buildings of hospitals and prisons as the material components 
(or as the 'form of content') and the discourses of medicine or criminology as 
the expressive components (or the 'form of expression'). See Gilles Deleuze, 
Foucault (Minneapolis, MN: University of Minnesota Press, 1988), p. 62. 

17. Jeffrey Pfeffer and Gerald R. Salancik, The External Control of Organizations. A 
Resource Dependence Perspective (Stanford, CA: Stanford University Press, 
2003), p. 46. 

18. Ibid., pp. 48-50. 

19. Ibid., p. 51. Despite these useful insights there is a major shortcoming to the 
theory of resource dependence. The authors' reliance on social constructi¬ 
vism to think about the way in which an organization (or rather, its 
administrative staff) perceives’ its relations with other organizations leads 



them to the idealist conclusion that an organization's environment is created 
by those relations to which it actually pays attention. As they put it: Noting 
that an organization's environment is enacted, or created by attentional 
processes, tends to shift the focus from characteristics of the objective 
environment to characteristics of the decision process by which organiza¬ 
tions select and ignore information' (ibid., p. 74). But why would anyone 
want to shift attention away from the objective distribution of opportunities 
and risks that an environment affords an organization? It is only by 
preserving the distinction between real opportunities to acquire resources 
(or real risks of losing autonomy) and the awareness that an organization 
may or may not have of them, that one can speak of 'missed opportunities' 
(or of 'underestimated risks') and of the effects that such mistaken 
evaluations may have on an organization's ability to cope with real 
dependencies. The notion of an 'enacted environment' is, in fact, quite 
useless, but the fact that the social constructivist part of the theory of 
resource dependence can be so easily separated from the rest shows that its 
role is mostly ceremonial rather than technical. 

20. Ibid., Ch. 6. 

21. Walter W. Powell, 'Neither market nor Hierarchy. Network forms of 
organization', in Michael Handel (ed.). The Sociology of Organizations 
(Thousand Oaks, CA: Sage, 2003), p. 326. 

22. John R. Munkirs and James I. Sturgeon, 'Oligopolistic cooperation: 
conceptual and empirical evidence of market structure evolution', in Marc 
R. Tool and Warren J. Samuels (eds). The Economy as a System of Power (New 
Brunswick, NJ: Transaction Press, 1989). 

23. Paul M. Hohenberg and Lynn Hollen Lees, The Making of Urban Europe 1000- 
1950 (Cambridge, MA: Harvard University Press, 1985), p. 202. 

24. Michael Best, The New Competition (Cambridge, MA: Harvard University 
Press, 1990), pp. 14-15. 

25. Ibid., p. 205. 

26. Annalee Saxenian, Regional Advantage. Culture and Competition in Silicon Valley 
and Route 128 (Cambridge, MA: Harvard University Press, 1996), pp. 2-3. 

27. Pfeffer and Salancik, The External Control of Organizations, pp. 94-5. 

28. Best, The New Competition, pp. 239-40. 

29. Howard T. Odum and Elizabeth C. Odum, Energy Basis for Man and Nature 
(New York: McGraw-Hill, 1981), p. 41. 

30. Saxenian, Regional Advantage, pp. 34-6. 

31. Pfeffer and Salancik, The External Control of Organizations, pp. 178-9. 

32. Walter W. Powell and Paul J. DiMaggio 'The iron cage revisited: institutional 
isomorphism and collective rationality in organizational fields', in Powell 
and DiMaggio (eds). The New Institutionalism in Organizational Analysis, pp. 



33. Michael T. Hannan and John Freeman, Organizational Ecology (Cambridge, 
MA: Harvard University Press, 1989), p. 66. 

34. Oliver E. Williamson, 'Transaction cost economics and organization theory', 
in Oliver E. Williamson (ed.). Organization Theory (New York: Oxford 
University Press, 1995), p. 223. 

35. Oliver E. Williamson, 'Chester Barnard and the incipient science of 
organization', ibid., p. 196. 

The focus of neo-institutional economists is at times too narrow (the only 
choice being between making or buying, or between internal hierarchies 
and external markets) so it does not cover all the possible resource 
interdependencies that may arise. In particular, division of labour among 
organizations of similar size (that is, in the absence of clear-cut domination 
by a much larger firm) may lead to specialization on products or activities 
which are dissimilar but closely complementary. This, in turn, presents firms 
with another choice, not to make or buy but to make or cooperate. The 
resulting interdependencies may lead to alliances or partnerships based on 
contracts for the transfer, exchange or pooling of technologies, standards 
and even personnel. See G.B. Richardson, 'The organization of industry', in 
Peter J. Buckley and Jonathan Michie (eds). Firms, Organizations and 
Contracts (Oxford: Oxford University Press, 2001), pp. 59-63. 

36. Terry M. Moe, 'The politics of structural choice: toward a theory of public 
bureaucracy', in Organization Theory, p. 125. 

37. Best, The New Competition, p. 82. 

38. Jeffrey L. Pressman and Aaron Wildavsky, Implementation (Berkeley, CA: 
University of California Press, 1984). Ch. 5. 

39. Daniel A. Mazmanian and Paul A. Sabatier, Implementation and Public Policy 
(Lanham, MD: University Press of America, 1989), p. 9. 

40. B. Dan Wood and Richard W. Waterman, Bureaucratic Dynamics (Boulder, 
CO: Westview Press, 1994), pp. 22-30. 

41. Pfeffer and Salancik, The External Control of Organizations, pp. 210-11. 

42. Charles Tilly, Stories. Identities, and Political Change (Lanham, MD: Rowman & 
Littlefield. 2002), p. 13. 

43. Hannu Nurmi, Comparing Voting Systems (Dordrecht: D. Reidel, 1987), pp. 2-3. 

44. James O. Freedman, Crisis and legitimacy. The Administrative Process and 
American Government (Cambridge: Cambridge University Press, 1978), pp. 

45. Ibid., pp. 44-6. 

46. Ibid., pp. 129-30 and 161-76. 

47. Wood and Waterman, Bureaucratic Dynamics pp. 33-7. 

48. Ibid., p. 144. 

49. Rolf Torstendahl, Bureaucratization in Northwestern Europe, 1880-1981 
(London: Routledge, 1991), pp. 203-16. 


50. David Sanders, Patterns of Political Instability (London: Macmillan. 1981), pp. 

51. While some discrepancy between statutory objectives and policy 
decisions is almost inevitable (if for no other reason than disagreements 
about how general rules apply to specific cases), such differences can be 
reduced if the statute stipulates unambiguous objectives, assigns 
implementation to sympathetic agencies who will give it high priority, 
minimizes the number of veto points and provides sufficient incentives 
(such as subsidies or compensatory changes in unrelated policies) to 
overcome recalcitrant officials, provides sufficient financial resources to 
conduct the technical analyses and process individual cases, and biases 
the decision rules and access points in favour of programme objectives. 
(Mazmanian and Sabatier, Implementation and Public Policy, p. 36) 

52. Douglass C. North, Institutions, Institutional Change and Economic Perf*rmance 
(New York: Cambridge University Press, 1995), pp. 120-31. 

53. Tilly. Stories, Identities, and Political Change, p. 129. 

54. T.K. Oommen, Citizenship, Nationality, and Ethnicity (Cambridge: Polity Press, 
1997). See p. 34 for the difference between state-led and state-seeking 
nationalisms and pp. 13 5-45 for mixtures in concrete cases. 

55. Charles Tilly, Big Structures, Large Processes, Huge Comparisons (New York: 
Russell Sage Foundation, 1984), pp. 103-11. 

Chapter 5 

1. Robert E. Park, 'The city: suggestions for investigation of human behaviour 
in the urban environment', in Robert E. Park and Ernest W. Burgess (eds), 
The City (Chicago, 1L: University of Chicago Press, 1984), pp. 4-6. 

2. Anthony Giddens, The Constitution of Society (Berkeley, CA: University of 
California Press, 1986), pp. 118-19. Giddens’ treatment of regionalized 
locales is similar to Deleuze and Guattari’s concept of a territory: a concept 
they develop in relation to animal territories but that is not confined to this 
example. To see the parallel, we must add to Giddens" definition in terms of 
rhythmic or periodic routines the expressive marking of boundaries. A 
territory is, in this sense, 'an act of rhythm that has become expressive'. 

Cf. Gilles Deleuze and Felix Guattari, A Thousand Plateaus (Minneapolis, 
MN: University of Minnesota Press, 1987), p. 315. Actually, there are three 
elements in the definition of a territorial assemblage. One needs 'a block of 
space-time constituted by the periodic repetition of [a] component' (ibid., p. 
313) made into a territory by marking its boundaries, drawing a circle 
around that uncertain and fragile centre, to organize a limited space' (ibid., 
p. 311). And, in addition to rhythm and boundary, there must be the 
possibility of opening up the circle, of venturing away from home through a 



gap in the border. This, of course, corresponds to the processes of 
deterritorialization which can open up an assemblage to future possibilities 
or even change its identity. 

3. Fernand Braudel, The Structures of Everyday Life (Berkeley, CA: University of 
California Press, 1992), p. 267. 

4. James E. Vance Jr. The Continuing City. Urban Morphology in Western Civilization 
(Baltimore, MD: Johns Hopkins University Press, 1990), pp. 24-5. 

5. Braudel, The Structures of Everyday Life, p. 308. 

6. Vance Jr, The Continuing City, p. 416. 

7. Ibid., p. 378. 

8. Braudel, The Structures of Everyday Life, p. 310. 

9. [bid., p. 317. 

10. (bid., p. 324. 

11. Michel Foucault, Discipline and Punish. The Birth of Prison (New York: Vintage 
Books, 1979), p. 172. 

12. Giddens, The Constitution of Society, p. 152. 

13. Vance Jr, The Continuing City, p. 175. 

14. Ibid., pp. 120 and 184-5. 'The central morphological truth learned in the 
bastides was that inter accessible and proportionate layout of the town is one 
of the more concrete expressions of functional equality, and a strong 
bulwark in its defense' (ibid., p. 200). 

1 5. Ibid., pp. 36-7. 

16. As the economist Thomas Shelling has shown, the dynamics behind these 
processes are those of people responding to an environment which consists 
of people responding to each other: given a group of people's preferences to 
live in proximity to similar groups, each decision made to move into or out 
of a neighbourhood will change the neighbourhood itself, influencing the 
future decisions of current residents and of people wanting residence there. 
See Thomas C. Schelling, Micromotives and Macrobehaviour (New York: 
Norton, 1978), Ch. 4. 

17. Robert E. Park, 'The city', in Park and Burgess, The City, p. 9. 

18. Vance Jr, The Continuing City, p. 316. 

19. Ernest W. Burgess, 'The growth of the city', in Park and Burgess, The City, p. 

20. Paul M. Hohenberg and Lynn Hollen I.ees, The Making of Urban Europe 1000 - 
1950 (Cambridge. MA: Harvard University Press, 1985), p. 299. 

21. Vance Jr, The Continuing City, p. 409. 

22. Ibid., pp. 412-13. 

23. Ibid., pp. 74-7. 

24. Braudel, The Structures of Everyday Life, pp. 484-9. 

25. Ibid., p. 486. 

26. Masahisa Fujita. Paul Krugman and Anthony J. Venables, The Spatial 

Economy. Cities, Regions, and International Trade (Cambridge, MA: MIT Press, 

1999), p. 4. See also Peter M. Allen, Cities and Regions as Self-Organizing 
Systems (Amsterdam: Gordon & Breach, 1997), p. 27. 

27. Vance Jr, The Continuing City, p. 373. 

28. Deleuze and Guattari view rhythmically repealed motifs and the counter¬ 
points they create with the external milieu as the two ways in which 
expressive components self-organize in territorial assemblages, including 
animal assemblages, transforming what was mere signature into a style. See 
Deleuze and Guattari, A Thousand Plateaus, p. 317. 

29. Spiro Kostoff, The City Shaped. Urban Patterns and Meanings throughout History 
(London: Bullfinch Press, 1991), pp. 284-5. 

30. Vance Jr, The Continuing City, p. 56. 

31. Braudel, The Structures of Everyday Life, p. 512. 

32. Vance Jr, The Continuing City, pp. 502-4. 

33. Hohenberg and Hollen Lees, The Making of Urban Europe, pp. 20-23 (for the 
period between the years 1000 and 1300);pp. 106-7 (1 500-1800); and pp. 
217-220 (1800-1900). 

34. Fujita et al., The Spatial Economy, p. 34. 

35. Allen, Cities and Regions as Self-Organizing Systems, p, 53. 

36. Hohenberg and Hollen Lees, The Making of Urban Europe, pp. 5 1-4. 

37. Ibid., p. 240. 

38. Fernand Braudel. The Perspective of the World (New York; Harper & Row, 
1979), pp. 27-31. 

39. Hohenberg and Hollen Lees, The Making of Urban Europe, p. 66. 

40. Braudel, The Perspective of the World, pp. 30-31. 

41. Hohenberg and Hollen Lees, The Making of Urban Europe, p. 6. 

42. Ibid., p. 281. 

43. Paul Kennedy, The Rise and Fall of the Great Powers. Economic Change and 
Military Conflict from 1500 to 2000 (New York: Random H*use, 1987), pp. 70- 

44. J. Craig Barker, International Law and International Relations (London: 
Continuum, 2000), pp. 5-8. For the five-year negotiation period see 
Geoffrey Parker, The Thirty Years' War (London: Routledge & Kegan Paul, 
1987), pp. 170-78. 

45. Kennedy, The Rise and Fall of the Great Powers, p. 86 (emphasis in the 

46. Fernand Braudel, The Wheels of Commerce (New York: Harper & Row, 1979), 
pp. 544-5. 

47. Ibid., p. 525 (my emphasis). 

48. Braudel, The Structures of Everyday Life, p. 527. 

49. Hohenberg and Hollen Lees, The Making of Urban Europe, p. 242. 

50. Kostoff, The City Shaped, pp. 2 11-1 5. 


51. Ibid., p. 217. 

52. I attempted to synthesize all available materials on the political history of 
languages and dialects in Manuel DeLanda, A Thousand Years of Nonlinear 
History (New York: Zone Books, 1 997), Ch. 3. 

53. Christopher Duffy, The Fortress in the Age of Vauban and Frederick the Great 
(London: Roulledge & Kegan Paul, 1985), p. 87. 

54. Peter J. Taylor. Political Geography (New York: Longman, 1985), pp. 113-15. 

55. Braudel introduced the term 'world-economy' to discuss the Mediterranean 
as a coherent economic area in Fernand Braudel, The Mediterranean. And the 
Mediterranean World in the Aye of Philip II, Vol. I. (Berkeley. CA: University of 
California Press, 1995), p. 419. Braudel attributes the original concept totwo 
German scholars in Braudel. The Perspective of the World, p. 634, n. 4. 

56. Immanuel Wallerstein, World-Systems Analysis. An Introduction (Durham. NC: 
Duke University Press, 2004), pp. 11-17. 

57. Ibid., p. 16. Wallerstein's macro-red unionism derives directly from his use 
of Hegelian totalities to conceptualize large-scale social entities. See 
Immanuel Wallerstein, The Capitalist World-Economy (Cambridge: Cambridge 
University Press, 1993), p. 4. 

58. Braudel, The Wheels of Commerce, p. 458. 


Arislotie 26-9 

Coding/Decoding 15-16, 19, 28, 59, 67, 

Diagram 31,125(16) 

Expressive Components 12, 16, 22, 50, 
54-5, 57, 60, 70-1, 88-9, 92, 97-8, 
100, 105-6, 115-16 
Material Components 12, 22,49- 50, 

5 3-4, 57, 60-1, 72-3, 81, 89, 91-2, 
95-7, 99, 105, 110, 114-15 
Terri torialization/ 

Deterri totalization 12, 16, 19, 28, 

50, 54-5, 58, 61-2, 67, 73^1, 81-3, 
89-90, 92 3, 98-9, 101-3, 106-7, 

Theory 3-5. 10-19, 21, 28, 30, 

33-4, 38-40, 121 (t 9, 10), 122 (f 13), 
123 (f 21), 130 (f 6) 

Bourdieu, Pierre 63-6 

Braudel, Fernand 17-18, 95-7, 106, 110, 

1 14-15, 1 18 
Bunge, Mnrio 10, 20-1 

Capacities 7, 10-11. 12, 17, 29, 35, 37-8, 
75, 89 

Cities 5, 34, 37 9, 41, 94, 103-12, 115-16 
Coleman, James 70 

Conversations 12, 16, 3 3, 52-5, 87 

Deleuze, Gilles 3-4, 10-11, 1 4, 47. 49 

Emergence 4. 6, 10, 14, 17, 32. 34, 38-9. 

47, 57, 70. 108, 129 (f 19) 
Enforcement 65, 68, 72, 80, 84, 89 
Essence 4, 26-31. 45, 48, 132 (f 35) 

by Causes 1 9-24, 31, 36-4 1 
by Reasons and Motives 22-4, 39-40, 
49, 57, 64, 95 

by Topological Constraints 31 

Foucault, Michel 72-3, 75, 99 

Giddens, Anthony 5, 9-10. 94-5, 99 
Goffrnan, Erving 5, 52-5, 119 
Government 33 4,36, 59, 67, 84-93, 

11 3 

Hacking, lan 2 
Hegel, G. W. F. 9-10 
Hume, David 47-51 

Identity 4, 10. 12, 14-15, 26, 28, 33, 

48-50, 54, 59, 87, 89-90, 102-3, 106, 
1 16 

Implementation 43, 81, 85 




Institutional Organizations 5-6, 12-13, 

15, 33-5, 37-9, 41-4, 66, 68-75, 77-9, 
83-4. 88-90, 113-14 
Intensity 7, 48, 52, 53, 55. 56, 107 
Interpersonal Networks 5-6, 12-13, 33-5, 
43-4, 56-9, 66 

Kennedy, Paul 112-13 

Language 2-3, 12, 14-16, 23, 44, 51, 55, 

58, 62, 66, 74-5, 81, 83, 91, 116, 122 
(f 14) 

Legitimacy 13, 68-71, 84, 86, 88-9 

Markets 17-18, 30, 32. 36, 115 
Micro-Macro Problem 4-6, 17, 32, 127 
(f 11), 128 (f 13) 

Ontology 1-2, 7, 28, 40, 126 (f 7 ) 

Populations 16-17, 21, 24-5, 34, 39, 41, 

59, 75, 99, 107 

Properties 6-7, 10-11,17, 27-9, 37, 56, 76 

of Exteriority 10-1 1, 16, 18, 45, 47-8, 
53, 56, 59, 63, 75-6 

of Interiority 9-11, 16, 18, 20, 45, 132 
(f 35) 

of Part-to-Whole 15, 27, 30, 33-40 

Resources 34-6, 42, 63-5, 70, 76-8, 104, 

Seamless Web 4, 9-10, 19, 21 
Significance 22-3, 55, 62, 80 

Individual 27-8, 30-1, 40, 48 
Universal 29-31. 40 
Social Classes 62-7 

Social Construction 3-4, 45, 62, 66, 133 
(f 47), 134 (( 8, 19) 

Social Movements 33-4, 36, 59-62 
Solidarity 13, 57, 80 
Subjectivity 32-3, 47-52 

Temporality 40-4 

Territorial State 93, 111-13, 114-18 

Tilly, Charles 5, 58-63, 66, 87, 92 

Vance, James 97,100-1 

Weber, Max 5, 22- 4, 30, 68-9, 74. 88,